Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,8 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+Data/inha_boundary_map.png filter=lfs diff=lfs merge=lfs -text
+Data/inha_boundary_polygon.png filter=lfs diff=lfs merge=lfs -text
+Data/inha_boundary_satellite.png filter=lfs diff=lfs merge=lfs -text
+데이터[[:space:]]수집/inha_boundary_with_points.png filter=lfs diff=lfs merge=lfs -text
+모수탐정단_제안서.pdf filter=lfs diff=lfs merge=lfs -text

.gradio/certificate.pem

@@ -0,0 +1,31 @@
+-----BEGIN CERTIFICATE-----
+MIIFazCCA1OgAwIBAgIRAIIQz7DSQONZRGPgu2OCiwAwDQYJKoZIhvcNAQELBQAw
+TzELMAkGA1UEBhMCVVMxKTAnBgNVBAoTIEludGVybmV0IFNlY3VyaXR5IFJlc2Vh
+cmNoIEdyb3VwMRUwEwYDVQQDEwxJU1JHIFJvb3QgWDEwHhcNMTUwNjA0MTEwNDM4
+WhcNMzUwNjA0MTEwNDM4WjBPMQswCQYDVQQGEwJVUzEpMCcGA1UEChMgSW50ZXJu
+ZXQgU2VjdXJpdHkgUmVzZWFyY2ggR3JvdXAxFTATBgNVBAMTDElTUkcgUm9vdCBY
+MTCCAiIwDQYJKoZIhvcNAQEBBQADggIPADCCAgoCggIBAK3oJHP0FDfzm54rVygc
+h77ct984kIxuPOZXoHj3dcKi/vVqbvYATyjb3miGbESTtrFj/RQSa78f0uoxmyF+
+0TM8ukj13Xnfs7j/EvEhmkvBioZxaUpmZmyPfjxwv60pIgbz5MDmgK7iS4+3mX6U
+A5/TR5d8mUgjU+g4rk8Kb4Mu0UlXjIB0ttov0DiNewNwIRt18jA8+o+u3dpjq+sW
+T8KOEUt+zwvo/7V3LvSye0rgTBIlDHCNAymg4VMk7BPZ7hm/ELNKjD+Jo2FR3qyH
+B5T0Y3HsLuJvW5iB4YlcNHlsdu87kGJ55tukmi8mxdAQ4Q7e2RCOFvu396j3x+UC
+B5iPNgiV5+I3lg02dZ77DnKxHZu8A/lJBdiB3QW0KtZB6awBdpUKD9jf1b0SHzUv
+KBds0pjBqAlkd25HN7rOrFleaJ1/ctaJxQZBKT5ZPt0m9STJEadao0xAH0ahmbWn
+OlFuhjuefXKnEgV4We0+UXgVCwOPjdAvBbI+e0ocS3MFEvzG6uBQE3xDk3SzynTn
+jh8BCNAw1FtxNrQHusEwMFxIt4I7mKZ9YIqioymCzLq9gwQbooMDQaHWBfEbwrbw
+qHyGO0aoSCqI3Haadr8faqU9GY/rOPNk3sgrDQoo//fb4hVC1CLQJ13hef4Y53CI
+rU7m2Ys6xt0nUW7/vGT1M0NPAgMBAAGjQjBAMA4GA1UdDwEB/wQEAwIBBjAPBgNV
+HRMBAf8EBTADAQH/MB0GA1UdDgQWBBR5tFnme7bl5AFzgAiIyBpY9umbbjANBgkq
+hkiG9w0BAQsFAAOCAgEAVR9YqbyyqFDQDLHYGmkgJykIrGF1XIpu+ILlaS/V9lZL
+ubhzEFnTIZd+50xx+7LSYK05qAvqFyFWhfFQDlnrzuBZ6brJFe+GnY+EgPbk6ZGQ
+3BebYhtF8GaV0nxvwuo77x/Py9auJ/GpsMiu/X1+mvoiBOv/2X/qkSsisRcOj/KK
+NFtY2PwByVS5uCbMiogziUwthDyC3+6WVwW6LLv3xLfHTjuCvjHIInNzktHCgKQ5
+ORAzI4JMPJ+GslWYHb4phowim57iaztXOoJwTdwJx4nLCgdNbOhdjsnvzqvHu7Ur
+TkXWStAmzOVyyghqpZXjFaH3pO3JLF+l+/+sKAIuvtd7u+Nxe5AW0wdeRlN8NwdC
+jNPElpzVmbUq4JUagEiuTDkHzsxHpFKVK7q4+63SM1N95R1NbdWhscdCb+ZAJzVc
+oyi3B43njTOQ5yOf+1CceWxG1bQVs5ZufpsMljq4Ui0/1lvh+wjChP4kqKOJ2qxq
+4RgqsahDYVvTH9w7jXbyLeiNdd8XM2w9U/t7y0Ff/9yi0GE44Za4rF2LN9d11TPA
+mRGunUHBcnWEvgJBQl9nJEiU0Zsnvgc/ubhPgXRR4Xq37Z0j4r7g1SgEEzwxA57d
+emyPxgcYxn/eR44/KJ4EBs+lVDR3veyJm+kXQ99b21/+jh5Xos1AnX5iItreGCc=
+-----END CERTIFICATE-----

Data/inha_boundary.cpg

@@ -0,0 +1 @@
+UTF-8

Data/inha_boundary.geojson

@@ -0,0 +1,172 @@
+{
+  "type": "FeatureCollection",
+  "features": [
+    {
+      "type": "Feature",
+      "geometry": {
+        "type": "Polygon",
+        "coordinates": [
+          [
+            [
+              126.6504913,
+              37.4488536
+            ],
+            [
+              126.6502018,
+              37.4490381
+            ],
+            [
+              126.6501274,
+              37.4490101
+            ],
+            [
+              126.650006,
+              37.449035
+            ],
+            [
+              126.6499295,
+              37.4493095
+            ],
+            [
+              126.6497967,
+              37.4495766
+            ],
+            [
+              126.6496642,
+              37.4498464
+            ],
+            [
+              126.6493324,
+              37.4504076
+            ],
+            [
+              126.6490067,
+              37.4509487
+            ],
+            [
+              126.649781,
+              37.4514003
+            ],
+            [
+              126.6495914,
+              37.4517312
+            ],
+            [
+              126.6501326,
+              37.45175
+            ],
+            [
+              126.6505999,
+              37.4517726
+            ],
+            [
+              126.6510982,
+              37.4518266
+            ],
+            [
+              126.6512838,
+              37.4518467
+            ],
+            [
+              126.6517828,
+              37.4519342
+            ],
+            [
+              126.6520048,
+              37.4519731
+            ],
+            [
+              126.6520885,
+              37.4519878
+            ],
+            [
+              126.6524237,
+              37.4520465
+            ],
+            [
+              126.6525693,
+              37.452072
+            ],
+            [
+              126.6528855,
+              37.4521371
+            ],
+            [
+              126.6530797,
+              37.4521045
+            ],
+            [
+              126.6536373,
+              37.4518765
+            ],
+            [
+              126.6544021,
+              37.4516675
+            ],
+            [
+              126.6564271,
+              37.4511471
+            ],
+            [
+              126.6578465,
+              37.4507776
+            ],
+            [
+              126.6569217,
+              37.4492734
+            ],
+            [
+              126.6553859,
+              37.4466845
+            ],
+            [
+              126.6547144,
+              37.4468884
+            ],
+            [
+              126.6545294,
+              37.4469713
+            ],
+            [
+              126.6544155,
+              37.4470167
+            ],
+            [
+              126.653386,
+              37.4474272
+            ],
+            [
+              126.6527874,
+              37.4476634
+            ],
+            [
+              126.6507639,
+              37.4484888
+            ],
+            [
+              126.6505808,
+              37.4485626
+            ],
+            [
+              126.6505271,
+              37.4487565
+            ],
+            [
+              126.6504913,
+              37.4488536
+            ]
+          ]
+        ]
+      },
+      "properties": {
+        "amenity": "university",
+        "education": "university",
+        "name": "인하대학교",
+        "name:en": "Inha University",
+        "name:ja": "仁荷大学校",
+        "name:zh": "仁荷大学",
+        "wikidata": "Q488851"
+      }
+    }
+  ]
+}

Data/inha_boundary.prj

@@ -0,0 +1 @@
+GEOGCS["GCS_WGS_1984",DATUM["D_WGS_1984",SPHEROID["WGS_1984",6378137.0,298.257223563]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]]

Data/zones.geojson

@@ -0,0 +1,7 @@
+{"type":"FeatureCollection","features":[{"type":"Feature","properties":{},
+"geometry":{"type":"Polygon","coordinates":[[[126.653513,37.450119],[126.6545,37.451635],[126.652805,37.452087],[126.649843,37.451652],[126.653513,37.450119]]]}},{"type":"Feature","properties":{},
+"geometry":{"type":"Polygon","coordinates":[[[126.653523,37.450059],[126.65009,37.451431],[126.649232,37.45092],[126.650208,37.449157],[126.650594,37.448807],[126.650659,37.44856],[126.652247,37.447921],[126.653523,37.450059]]]}},{"type":"Feature","properties":{},
+"geometry":{"type":"Polygon","coordinates":[[[126.653588,37.450119],[126.654553,37.451618],[126.656442,37.451116],[126.655476,37.449369],[126.653588,37.450119]]]}},{"type":"Feature","properties":{},
+"geometry":{"type":"Polygon","coordinates":[[[126.65554,37.449327],[126.656474,37.451073],[126.657826,37.450766],[126.656656,37.448875],[126.65554,37.449327]]]}},{"type":"Feature","properties":{},
+"geometry":{"type":"Polygon","coordinates":[[[126.655465,37.449284],[126.653534,37.450021],[126.6523,37.447904],[126.65436,37.447053],[126.655465,37.449284]]]}},{"type":"Feature","properties":{},
+"geometry":{"type":"Polygon","coordinates":[[[126.655519,37.449276],[126.654371,37.447019],[126.655374,37.446703],[126.656662,37.448863],[126.655519,37.449276]]]}}]}

Data/zones_최종.geojson

@@ -0,0 +1 @@
+{"type":"FeatureCollection","features":[{"type":"Feature","properties":{},"geometry":{"type":"Polygon","coordinates":[[[126.649822,37.451397],[126.65031,37.451388],[126.653615,37.450085],[126.65215,37.447926],[126.650605,37.44859],[126.650509,37.448871],[126.650208,37.449071],[126.650047,37.449059],[126.64905,37.450941],[126.649822,37.451397]]]}},{"type":"Feature","properties":{},"geometry":{"type":"Polygon","coordinates":[[[126.653631,37.450081],[126.65215,37.447926],[126.654484,37.447002],[126.655846,37.449184],[126.653631,37.450081]]]}},{"type":"Feature","properties":{},"geometry":{"type":"Polygon","coordinates":[[[126.654478,37.446985],[126.655852,37.449191],[126.656678,37.448875],[126.655374,37.446695],[126.654478,37.446985]]]}},{"type":"Feature","properties":{},"geometry":{"type":"Polygon","coordinates":[[[126.656688,37.448875],[126.655401,37.449382],[126.655433,37.449676],[126.656442,37.451145],[126.657842,37.450775],[126.656688,37.448875]]]}},{"type":"Feature","properties":{},"geometry":{"type":"Polygon","coordinates":[[[126.655406,37.449378],[126.653529,37.450145],[126.654623,37.45161],[126.656447,37.451145],[126.655417,37.449702],[126.655406,37.449378]]]}},{"type":"Feature","properties":{},"geometry":{"type":"Polygon","coordinates":[[[126.653513,37.45014],[126.650316,37.451409],[126.649806,37.451409],[126.649624,37.451707],[126.650686,37.451763],[126.65288,37.452104],[126.654607,37.451601],[126.653513,37.45014]]]}}]}

Data/측정결과_수정중.csv

@@ -0,0 +1,113 @@
+lat,lon,dB
+37.4516638,126.6527864,2.0
+37.4517164,126.6525593,30.0
+37.4512905,126.653364,3.0
+37.4514279,126.652793,4.0
+37.4511085,126.6535225,5.0
+37.4510701,126.6531148,6.0
+37.4513171,126.652632,7.0
+37.4513746,126.6517523,24.0
+37.4512256,126.6514948,21.0
+37.4510425,126.6515538,22.0
+37.4510595,126.6526267,8.0
+37.4505606,126.6533733,14.0
+37.4506117,126.6531748,25.0
+37.4514975,126.6537273,28.0
+37.4509056,126.6523755,26.0
+37.450863,126.6535503,11.0
+37.4517914,126.6530514,27.0
+37.4514336,126.6522414,31.0
+37.451127,126.6513348,23.0
+37.4505351,126.6524077,18.0
+37.4507437,126.6524077,16.0
+37.4503988,126.6532875,19.0
+37.4510631,126.6536844,10.0
+37.4515699,126.6534377,29.0
+37.4506799,126.6527081,17.0
+37.4512906,126.6521779,32.0
+37.4509074,126.6521297,15.0
+37.4510309,126.6528753,9.0
+37.4512012,126.6539697,13.0
+37.4507669,126.6537015,12.0
+37.4509659,126.6549157,70.0
+37.4508935,126.6552698,71.0
+37.4508636,126.6558974,69.0
+37.4511234,126.6544705,68.0
+37.4513577,126.6545939,66.0
+37.4505187,126.6540735,67.0
+37.4505826,126.6550338,72.0
+37.4506465,126.6546851,73.0
+37.4498331,126.6546207,90.0
+37.4505016,126.6553824,78.0
+37.4503569,126.6549479,74.0
+37.4500523,126.6554818,79.0
+37.4498416,126.6553664,80.0
+37.449931,126.6551249,84.0
+37.4501013,126.6545617,86.0
+37.4500119,126.6549318,83.0
+37.450259,126.6542881,81.0
+37.4502206,126.6544383,82.0
+37.4501525,126.654213,87.0
+37.4503654,126.655495,77.0
+37.4497777,126.6547173,91.0
+37.4505655,126.655243,76.0
+37.4505698,126.6506188,41.0
+37.4503398,126.6505276,40.0
+37.4506251,126.6509514,42.0
+37.4504676,126.6517346,34.0
+37.4506252,126.6513859,33.0
+37.4499438,126.6512143,38.0
+37.4499054,126.6506027,39.0
+37.4500971,126.651917,37.0
+37.4504548,126.6550123,75.0
+37.4501865,126.6517293,36.0
+37.4502887,126.6514664,35.0
+37.4499437,126.6552644,85.0
+37.4495805,126.6526316,46.0
+37.4492398,126.6524492,48.0
+37.4497977,126.6530554,45.0
+37.4500532,126.6527818,43.0
+37.4499255,126.6529267,44.0
+37.4489332,126.6520201,49.0
+37.4489204,126.6522293,50.0
+37.449474,126.6528355,47.0
+37.4494917,126.6541855,57.0
+37.4499771,126.65462,88.0
+37.4497514,126.6551403,89.0
+37.4486569,126.6536919,53.0
+37.4489806,126.6539548,54.0
+37.4496024,126.654502,58.0
+37.4497003,126.6544966,60.0
+37.4499814,126.6539816,59.0
+37.4495939,126.654797,61.0
+37.4495172,126.6545878,56.0
+37.4493171,126.6552154,62.0
+37.4491893,126.6552101,63.0
+37.4490317,126.6551242,65.0
+37.4491978,126.655076,64.0
+37.4488827,126.6542498,55.0
+37.4479436,126.6529785,51.0
+37.4503967,126.6529463,20.0
+37.4486122,126.6533164,52.0
+37.448078,126.6556807,37.0
+37.4483203,126.6558354,41.0
+37.4483203,126.655504,47.0
+37.44746,126.655224,52.0
+37.4474539,126.6555703,45.0
+37.4476721,126.6550472,55.0
+37.4477023,126.655364,49.0
+37.4477751,126.6551283,56.0
+37.4488898,126.6557323,53.0
+37.4490655,126.6560712,55.0
+37.4489565,126.6559828,54.0
+37.4492533,126.656108,59.0
+37.4494472,126.6563216,63.0
+37.4497198,126.6561375,58.0
+37.4499379,126.656248,56.0
+37.4485566,126.6557102,59.0
+37.4483627,126.6551062,42.0
+37.4506589,126.6570141,51.0
+37.4503802,126.656469,58.0
+37.4505559,126.6571835,48.0
+37.4492957,126.656329,58.0
+37.4495078,126.6564248,61.0

Data/측정결과_최종.csv

@@ -0,0 +1,113 @@
+lat,lon,dB
+37.4516638,126.6527864,50.0
+37.4512905,126.653364,49.0
+37.4514279,126.652793,48.0
+37.4511085,126.6535225,49.0
+37.4510701,126.6531148,48.0
+37.4513171,126.652632,47.0
+37.4510595,126.6526267,50.0
+37.4510309,126.6528753,51.0
+37.4510631,126.6536844,57.0
+37.450863,126.6535503,52.0
+37.4507669,126.6537015,51.0
+37.4512012,126.6539697,47.0
+37.4505606,126.6533733,47.0
+37.4509074,126.6521297,58.0
+37.4507437,126.6524077,49.0
+37.4506799,126.6527081,48.0
+37.4505351,126.6524077,53.0
+37.4503988,126.6532875,47.0
+37.4503967,126.6529463,46.0
+37.4512256,126.6514948,47.0
+37.4510425,126.6515538,46.0
+37.451127,126.6513348,46.0
+37.4513746,126.6517523,42.0
+37.4506117,126.6531748,44.0
+37.4509056,126.6523755,42.0
+37.4517914,126.6530514,55.0
+37.4514975,126.6537273,42.0
+37.4515699,126.6534377,41.0
+37.4517164,126.6525593,39.0
+37.4514336,126.6522414,50.0
+37.4512906,126.6521779,42.0
+37.4506252,126.6513859,53.0
+37.4504676,126.6517346,45.0
+37.4502887,126.6514664,50.0
+37.4501865,126.6517293,47.0
+37.4500971,126.651917,49.0
+37.4499438,126.6512143,50.0
+37.4499054,126.6506027,51.0
+37.4503398,126.6505276,49.0
+37.4505698,126.6506188,43.0
+37.4506251,126.6509514,45.0
+37.4500532,126.6527818,46.0
+37.4499255,126.6529267,43.0
+37.4497977,126.6530554,44.0
+37.4495805,126.6526316,48.0
+37.449474,126.6528355,52.0
+37.4492398,126.6524492,53.0
+37.4489332,126.6520201,50.0
+37.4489204,126.6522293,46.0
+37.4479436,126.6529785,49.0
+37.4486122,126.6533164,46.0
+37.4486569,126.6536919,43.0
+37.4489806,126.6539548,44.0
+37.4488827,126.6542498,47.0
+37.4495172,126.6545878,38.0
+37.4494917,126.6541855,51.0
+37.4496024,126.654502,52.0
+37.4499814,126.6539816,52.0
+37.4497003,126.6544966,46.0
+37.4495939,126.654797,54.0
+37.4493171,126.6552154,46.0
+37.4491893,126.6552101,47.0
+37.4491978,126.655076,45.0
+37.4490317,126.6551242,47.0
+37.4513577,126.6545939,51.0
+37.4505187,126.6540735,53.0
+37.4511234,126.6544705,60.0
+37.4508636,126.6558974,43.0
+37.4509659,126.6549157,44.0
+37.4508935,126.6552698,39.0
+37.4505826,126.6550338,56.0
+37.4506465,126.6546851,52.0
+37.4503569,126.6549479,59.0
+37.4504548,126.6550123,54.0
+37.4505655,126.655243,42.0
+37.4503654,126.655495,47.0
+37.4505016,126.6553824,41.0
+37.4500523,126.6554818,41.0
+37.4498416,126.6553664,46.0
+37.450259,126.6542881,44.0
+37.4502206,126.6544383,43.0
+37.4500119,126.6549318,44.0
+37.449931,126.6551249,51.0
+37.4499437,126.6552644,46.0
+37.4501013,126.6545617,50.0
+37.4501525,126.654213,48.0
+37.4499771,126.65462,48.0
+37.4497514,126.6551403,52.0
+37.4498331,126.6546207,48.0
+37.4497777,126.6547173,52.0
+37.448078,126.6556807,37.0
+37.4483203,126.6558354,41.0
+37.4483203,126.655504,47.0
+37.44746,126.655224,52.0
+37.4474539,126.6555703,45.0
+37.4476721,126.6550472,55.0
+37.4477023,126.655364,49.0
+37.4477751,126.6551283,56.0
+37.4488898,126.6557323,53.0
+37.4490655,126.6560712,55.0
+37.4489565,126.6559828,54.0
+37.4492533,126.656108,59.0
+37.4494472,126.6563216,63.0
+37.4497198,126.6561375,58.0
+37.4499379,126.656248,56.0
+37.4485566,126.6557102,59.0
+37.4483627,126.6551062,42.0
+37.4506589,126.6570141,51.0
+37.4503802,126.656469,58.0
+37.4505559,126.6571835,48.0
+37.4492957,126.656329,58.0
+37.4495078,126.6564248,61.0

README.md

@@ -1,12 +1,6 @@
 ---
-title: Sampling Proj
-emoji: 🐨
-colorFrom: pink
-colorTo: blue
-sdk: gradio
-sdk_version: 6.0.1
+title: Sampling_Proj
 app_file: app.py
-pinned: false
+sdk: gradio
+sdk_version: 5.35.0
 ---
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,539 @@
+# sampling_sim_app.py
+import os
+import io
+
+import matplotlib
+matplotlib.use("Agg")
+
+import numpy as np
+import pandas as pd
+import geopandas as gpd
+import matplotlib.pyplot as plt
+from shapely.geometry import Point
+from shapely import wkt as shp_wkt
+from PIL import Image
+import contextily as ctx  # ← 추가
+
+import gradio as gr
+
+# ----------------- 경로 설정 -----------------
+BASE_DIR = os.path.dirname(__file__)
+DATA_DIR = os.path.join(BASE_DIR, "Data")
+
+SHP_PATH = os.path.join(DATA_DIR, "inha_boundary.shp")
+ZONES_PATH = os.path.join(DATA_DIR, "zones_최종.geojson")
+
+# 측정결과 CSV (lat, lon, dB 또는 위치, dB)
+CSV_PATH = os.path.join(DATA_DIR, "측정결과_최종.csv")
+
+
+# ---------- 유틸: Figure -> numpy ----------
+def fig_to_array(fig, dpi=100):
+    buf = io.BytesIO()
+    fig.savefig(buf, format="png", dpi=dpi, bbox_inches="tight")
+    buf.seek(0)
+    img = np.array(Image.open(buf))
+    plt.close(fig)
+    return img
+
+
+# ---------- 경계 SHP 로딩 ----------
+def load_boundary():
+    gdf4326 = gpd.read_file(SHP_PATH).to_crs(epsg=4326)
+    gdf3857 = gdf4326.to_crs(epsg=3857)
+
+    poly3857 = gdf3857.geometry.unary_union
+    if poly3857.geom_type == "MultiPolygon":
+        poly3857 = max(poly3857.geoms, key=lambda g: g.area)
+
+    return gdf4326, gdf3857, poly3857
+
+
+# ---------- 측정결과 CSV 로딩 ----------
+def load_points_from_csv():
+    try:
+        df_u = pd.read_csv(CSV_PATH, encoding="cp949")
+    except UnicodeDecodeError:
+        df_u = pd.read_csv(CSV_PATH, encoding="utf-8")
+
+    df_u.columns = [c.strip() for c in df_u.columns]
+    df_u = df_u.loc[:, ~df_u.columns.str.contains("^Unnamed")]
+
+    # 케이스 B: 위치, dB
+    if "위치" in df_u.columns and "dB" in df_u.columns:
+        pos_series = df_u["위치"].astype(str).str.strip()
+
+        def safe_load_wkt(s):
+            s_up = s.upper()
+            if "POINT" not in s_up:
+                return None
+            try:
+                return shp_wkt.loads(s)
+            except Exception:
+                return None
+
+        geom = pos_series.apply(safe_load_wkt)
+        mask = geom.notnull()
+        if not mask.any():
+            raise ValueError("'위치' 컬럼에 유효한 POINT WKT가 없다.")
+
+        geom_valid = geom[mask]
+        lons = geom_valid.apply(lambda g: round(g.x, 7))
+        lats = geom_valid.apply(lambda g: round(g.y, 7))
+        dBs = pd.to_numeric(df_u.loc[mask, "dB"], errors="coerce")
+
+        df = pd.DataFrame({"lat": lats, "lon": lons, "dB": dBs}).dropna(
+            subset=["lat", "lon"]
+        )
+        return df
+
+    # 케이스 A: lat, lon, dB
+    if {"lat", "lon", "dB"}.issubset(df_u.columns):
+        df = df_u[["lat", "lon", "dB"]].copy()
+        df["lat"] = pd.to_numeric(df["lat"], errors="coerce").round(7)
+        df["lon"] = pd.to_numeric(df["lon"], errors="coerce").round(7)
+        df["dB"] = pd.to_numeric(df["dB"], errors="coerce")
+        df = df.dropna(subset=["lat", "lon"])
+        return df
+
+    raise ValueError("CSV는 (lat, lon, dB) 또는 (위치, dB) 형식이어야 한다.")
+
+
+# ---------- 구역 GeoJSON 로딩 ----------
+def load_zones(poly3857):
+    zones4326 = gpd.read_file(ZONES_PATH).to_crs(epsg=4326)
+    zones3857 = zones4326.to_crs(epsg=3857)
+
+    if "zone_id" not in zones3857.columns:
+        zones3857["zone_id"] = np.arange(1, len(zones3857) + 1)
+
+    zones3857["geom_clip"] = zones3857.geometry.intersection(poly3857)
+    zones3857["area"] = zones3857["geom_clip"].area
+    zones3857 = zones3857[zones3857["area"] > 0].copy()
+
+    A_total = zones3857["area"].sum()
+    zones3857["W_area"] = zones3857["area"] / A_total
+
+    return zones4326, zones3857
+
+
+# ---------- 포인트에 zone_id 붙이기 ----------
+def attach_zone_id(points_df, zones3857):
+    pts4326 = gpd.GeoDataFrame(
+        points_df.copy(),
+        geometry=gpd.points_from_xy(points_df["lon"], points_df["lat"]),
+        crs=4326,
+    )
+    pts3857 = pts4326.to_crs(epsg=3857)
+
+    zones_clip = zones3857[["zone_id", "geom_clip"]].copy()
+    zones_clip = zones_clip.set_geometry("geom_clip")
+    zones_clip = zones_clip.rename_geometry("geometry")
+
+    joined = gpd.sjoin(
+        pts3857,
+        zones_clip[["zone_id", "geometry"]],
+        how="inner",
+        predicate="within",
+    )
+
+    out = joined[["lat", "lon", "dB", "zone_id"]].copy().reset_index(drop=True)
+    return out
+
+
+# ---------- 공통 준비: 경계/구역/포인트/층정보 ----------
+def prepare_sampling_data():
+    boundary4326, boundary3857, poly3857 = load_boundary()
+    zones4326, zones3857 = load_zones(poly3857)
+    df_points = load_points_from_csv()
+    points_zone = attach_zone_id(df_points, zones3857)
+
+    # 층별 모집단 크기 N_h
+    zone_sizes = points_zone.groupby("zone_id").size().rename("N_h")
+    zones_info = zones3857[["zone_id", "area", "W_area"]].merge(
+        zone_sizes, on="zone_id", how="left"
+    )
+    zones_info["N_h"] = zones_info["N_h"].fillna(0).astype(int)
+
+    # 층별 표준편차 S_h (네이만 배분용)
+    strata_std = points_zone.groupby("zone_id")["dB"].std(ddof=1).rename("S_h")
+    zones_info = zones_info.merge(strata_std, on="zone_id", how="left")
+    zones_info["S_h"] = zones_info["S_h"].fillna(0.0)
+
+    return boundary4326, zones4326, df_points, points_zone, zones_info
+
+
+# ---------- 면적비례 n_h 계산 ----------
+def allocate_n_by_area(zones_info, n_total):
+    zones = zones_info.copy()
+    zones["n_raw"] = n_total * zones["W_area"]
+    zones["n"] = zones["n_raw"].round().astype(int)
+
+    diff = n_total - zones["n"].sum()
+    if diff != 0:
+        order = zones["n_raw"].sort_values(ascending=(diff < 0)).index
+        for idx in order[:abs(diff)]:
+            zones.loc[idx, "n"] += 1 if diff > 0 else -1
+
+    zones["n"] = zones[["n", "N_h"]].min(axis=1)
+    return zones
+
+
+# ---------- 네이만 배분 n_h 계산 ----------
+def allocate_n_neyman(zones_info, n_total):
+    zones = zones_info.copy()
+    if "S_h" not in zones.columns:
+        raise ValueError("zones_info에 S_h가 없다. 네이만 배분 전에 S_h를 계산해야 한다.")
+
+    zones["S_h"] = zones["S_h"].fillna(0.0)
+
+    # 가중치 ∝ N_h * S_h
+    weight = zones["N_h"] * zones["S_h"]
+    if weight.sum() <= 0:
+        # 분산이 전부 0이면 N_h 비례로 배분
+        weight = zones["N_h"].copy()
+
+    zones["n_raw"] = n_total * weight / weight.sum()
+    zones["n"] = zones["n_raw"].round().astype(int)
+
+    diff = n_total - zones["n"].sum()
+    if diff != 0:
+        order = zones["n_raw"].sort_values(ascending=(diff < 0)).index
+        for idx in order[:abs(diff)]:
+            zones.loc[idx, "n"] += 1 if diff > 0 else -1
+
+    zones["n"] = zones[["n", "N_h"]].min(axis=1)
+    return zones
+
+
+# ---------- 시뮬레이션 ----------
+def simulate_sampling(points_zone_df, zones_info, n, R, strat_label):
+    y_all = points_zone_df["dB"].to_numpy()
+    N = len(y_all)
+    mu_pop = float(np.mean(y_all))
+
+    zones = zones_info.set_index("zone_id")
+    W_h = zones["N_h"] / zones["N_h"].sum()
+    n_h = zones["n"]
+
+    idx_by_zone = {}
+    for z in zones.index:
+        idx_by_zone[z] = points_zone_df.index[
+            points_zone_df["zone_id"] == z
+        ].to_numpy()
+
+    srs_means = []
+    strat_means = []
+
+    for _ in range(R):
+        # SRS
+        n_eff = min(n, N)
+        idx_srs = np.random.choice(N, size=n_eff, replace=False)
+        mu_srs = float(np.mean(y_all[idx_srs]))
+        srs_means.append(mu_srs)
+
+        # Stratified
+        mu_str = 0.0
+        for z in zones.index:
+            N_h = int(zones.loc[z, "N_h"])
+            nh = int(n_h.loc[z])
+            if N_h == 0 or nh == 0:
+                continue
+            idx_pool = idx_by_zone[z]
+            if len(idx_pool) < nh:
+                nh = len(idx_pool)
+            choose = np.random.choice(idx_pool, size=nh, replace=False)
+            y_h = points_zone_df.loc[choose, "dB"].to_numpy()
+            ybar_h = float(np.mean(y_h))
+            mu_str += float(W_h.loc[z]) * ybar_h
+        strat_means.append(mu_str)
+
+    srs_means = np.array(srs_means)
+    strat_means = np.array(strat_means)
+
+    def metrics(arr, name):
+        mean_hat = float(arr.mean())
+        var_hat = float(arr.var(ddof=1))
+        bias = mean_hat - mu_pop
+        mse = float(((arr - mu_pop) ** 2).mean())
+        return {
+            "design": name,
+            "mean_hat": mean_hat,
+            "var_hat": var_hat,
+            "bias": bias,
+            "MSE": mse,
+        }
+
+    m_srs = metrics(srs_means, "SRS")
+    m_str = metrics(strat_means, strat_label)
+
+    df_metrics = pd.DataFrame([m_srs, m_str])
+    df_metrics["deff_vs_SRS"] = df_metrics["var_hat"] / df_metrics["var_hat"].iloc[0]
+
+    # 숫자 4째 자리까지 반올림
+    num_cols = ["mean_hat", "var_hat", "bias", "MSE", "deff_vs_SRS"]
+    df_metrics[num_cols] = df_metrics[num_cols].round(4)
+
+    return mu_pop, srs_means, strat_means, df_metrics
+
+
+# ---------- 구역 요약 테이블 ----------
+def build_zones_table(zones_alloc):
+    zones_table = zones_alloc[["zone_id", "area", "W_area", "N_h", "n"]].copy()
+    zones_table = zones_table.rename(
+        columns={
+            "area": "area_m2",
+            "W_area": "area_weight",
+            "N_h": "N_pop",
+            "n": "n_alloc",
+        }
+    )
+    zones_table["sampling_frac"] = zones_table["n_alloc"] / zones_table["N_pop"]
+
+    total_area_m2 = zones_table["area_m2"].sum()
+    total_area_weight = zones_table["area_weight"].sum()
+    total_N = zones_table["N_pop"].sum()
+    total_n = zones_table["n_alloc"].sum()
+    total_sampling_frac = total_n / total_N if total_N > 0 else np.nan
+
+    total_row = {
+        "zone_id": "���계",
+        "area_m2": total_area_m2,
+        "area_weight": total_area_weight,
+        "N_pop": total_N,
+        "n_alloc": total_n,
+        "sampling_frac": total_sampling_frac,
+    }
+
+    zones_table = zones_table[
+        ["zone_id", "area_m2", "area_weight", "N_pop", "n_alloc", "sampling_frac"]
+    ]
+    zones_table = pd.concat([zones_table, pd.DataFrame([total_row])], ignore_index=True)
+
+    # 숫자 컬럼 4째 자리까지 반올림
+    num_cols_zone = ["area_m2", "area_weight", "N_pop", "n_alloc", "sampling_frac"]
+    for col in num_cols_zone:
+        zones_table[col] = pd.to_numeric(zones_table[col], errors="coerce").round(4)
+
+    return zones_table, float(total_N), float(total_n)
+
+
+# ---------- 시각화 ----------
+def plot_population_map(boundary4326, zones4326, points_df):
+    g_boundary = boundary4326.to_crs(epsg=3857)
+    g_zones = zones4326.to_crs(epsg=3857)
+    g_pts = gpd.GeoDataFrame(
+        points_df.copy(),
+        geometry=gpd.points_from_xy(points_df["lon"], points_df["lat"]),
+        crs=4326,
+    ).to_crs(epsg=3857)
+
+    fig, ax = plt.subplots(figsize=(8, 8))
+
+    # 베이스맵 먼저 깔기
+    minx, miny, maxx, maxy = g_boundary.total_bounds
+    ax.set_xlim(minx, maxx)
+    ax.set_ylim(miny, maxy)
+    ctx.add_basemap(ax, source=ctx.providers.OpenStreetMap.Mapnik, alpha=0.7)
+
+    # 그 위에 경계/구역/포인트
+    g_boundary.plot(ax=ax, facecolor="none", edgecolor="black", linewidth=2)
+    if len(g_zones) > 0:
+        g_zones.plot(
+            ax=ax,
+            facecolor="none",
+            edgecolor="red",
+            linewidth=1.5,
+            alpha=0.7,
+        )
+
+    g_pts.plot(
+        ax=ax,
+        column="dB",
+        cmap="viridis",
+        markersize=20,
+        edgecolor="black",
+        linewidth=0.2,
+        legend=False,
+        alpha=0.9,
+    )
+
+    ax.set_title("Population points with zones", fontsize=12)
+    ax.axis("off")
+    plt.tight_layout()
+    return fig_to_array(fig)
+
+
+def plot_histograms(srs_means, strat_means, mu_pop, strat_label):
+    fig, ax = plt.subplots(figsize=(8, 5))
+
+    ax.hist(srs_means, bins=20, alpha=0.5, label="SRS")
+    ax.hist(strat_means, bins=20, alpha=0.5, label=strat_label)
+    ax.axvline(mu_pop, color="k", linestyle="--", label="Population mean")
+
+    ax.set_xlabel("Sample mean (dB)")
+    ax.set_ylabel("Frequency")
+    ax.legend()
+    ax.set_title("Sampling distribution of mean")
+    plt.tight_layout()
+    return fig_to_array(fig)
+
+
+# ---------- Gradio 콜백: 면적비례 ----------
+def run_sim_area(n, R):
+    n = int(n)
+    R = int(R)
+
+    boundary4326, zones4326, df_points, points_zone, zones_info = prepare_sampling_data()
+
+    # 면적비례 배분
+    zones_alloc = allocate_n_by_area(zones_info, n)
+    zones_table, total_N, total_n = build_zones_table(zones_alloc)
+
+    mu_pop, srs_means, strat_means, df_metrics = simulate_sampling(
+        points_zone, zones_alloc, n, R, "Stratified (area-proportional)"
+    )
+
+    map_img = plot_population_map(boundary4326, zones4326, df_points)
+    hist_img = plot_histograms(
+        srs_means, strat_means, mu_pop, "Stratified (area-proportional)"
+    )
+
+    summary = (
+        f"[면적비례 층화]\n"
+        f"모집단 포인트 수: {int(total_N)}개\n"
+        f"구역 수: {len(zones_alloc)}개\n"
+        f"모집단 평균 dB: {mu_pop:.4f}\n"
+        f"총 표본크기 n = {int(total_n)} (입력값 {n}), 반복 R = {R}"
+    )
+
+    return map_img, hist_img, df_metrics, zones_table, summary
+
+
+# ---------- Gradio 콜백: 네이만 ----------
+def run_sim_neyman(n, R):
+    n = int(n)
+    R = int(R)
+
+    boundary4326, zones4326, df_points, points_zone, zones_info = prepare_sampling_data()
+
+    # 네이만 배분
+    zones_alloc = allocate_n_neyman(zones_info, n)
+    zones_table, total_N, total_n = build_zones_table(zones_alloc)
+
+    mu_pop, srs_means, strat_means, df_metrics = simulate_sampling(
+        points_zone, zones_alloc, n, R, "Stratified (Neyman)"
+    )
+
+    map_img = plot_population_map(boundary4326, zones4326, df_points)
+    hist_img = plot_histograms(
+        srs_means, strat_means, mu_pop, "Stratified (Neyman)"
+    )
+
+    summary = (
+        f"[네이만 층화]\n"
+        f"모집단 포인트 수: {int(total_N)}개\n"
+        f"구역 수: {len(zones_alloc)}개\n"
+        f"모집단 평균 dB: {mu_pop:.4f}\n"
+        f"총 표본크기 n = {int(total_n)} (입력값 {n}), 반복 R = {R}"
+    )
+
+    return map_img, hist_img, df_metrics, zones_table, summary
+
+
+# ---------- Gradio 앱 ----------
+def build_app():
+    with gr.Blocks(title="표본추출 시뮬레이션 (SRS vs 층화)") as demo:
+
+        with gr.Tab("SRS vs 면적비례 층화"):
+            with gr.Row():
+                n_area = gr.Number(label="표본크기 n", value=60, precision=0)
+                R_area = gr.Number(label="시뮬레이션 반복 횟수 R", value=500, precision=0)
+
+            run_btn_area = gr.Button("면적비례 층화 시뮬레이션 실행", variant="primary")
+
+            with gr.Row():
+                map_img_area = gr.Image(label="Population + zones")
+                hist_img_area = gr.Image(label="Sampling distribution")
+
+            metrics_table_area = gr.Dataframe(
+                headers=[
+                    "design",
+                    "mean_hat",
+                    "var_hat",
+                    "bias",
+                    "MSE",
+                    "deff_vs_SRS",
+                ],
+                label="성능 비교 지표",
+                interactive=False,
+            )
+
+            zones_table_area = gr.Dataframe(
+                label="구역별 면적 및 모집단/배정 표본수",
+                interactive=False,
+            )
+
+            summary_md_area = gr.Markdown()
+
+            run_btn_area.click(
+                fn=run_sim_area,
+                inputs=[n_area, R_area],
+                outputs=[
+                    map_img_area,
+                    hist_img_area,
+                    metrics_table_area,
+                    zones_table_area,
+                    summary_md_area,
+                ],
+            )
+
+        with gr.Tab("SRS vs 네이만 층화"):
+            with gr.Row():
+                n_neyman = gr.Number(label="표본크기 n", value=60, precision=0)
+                R_neyman = gr.Number(label="시뮬레이션 반복 횟수 R", value=500, precision=0)
+
+            run_btn_neyman = gr.Button("네이만 층화 시뮬레이션 실행", variant="primary")
+
+            with gr.Row():
+                map_img_neyman = gr.Image(label="Population + zones")
+                hist_img_neyman = gr.Image(label="Sampling distribution")
+
+            metrics_table_neyman = gr.Dataframe(
+                headers=[
+                    "design",
+                    "mean_hat",
+                    "var_hat",
+                    "bias",
+                    "MSE",
+                    "deff_vs_SRS",
+                ],
+                label="성능 비교 지표",
+                interactive=False,
+            )
+
+            zones_table_neyman = gr.Dataframe(
+                label="구역별 면적 및 모집단/배정 표본수",
+                interactive=False,
+            )
+
+            summary_md_neyman = gr.Markdown()
+
+            run_btn_neyman.click(
+                fn=run_sim_neyman,
+                inputs=[n_neyman, R_neyman],
+                outputs=[
+                    map_img_neyman,
+                    hist_img_neyman,
+                    metrics_table_neyman,
+                    zones_table_neyman,
+                    summary_md_neyman,
+                ],
+            )
+
+    return demo
+
+
+if __name__ == "__main__":
+    app = build_app()
+    app.launch(share=True)

draw_zones.html

@@ -0,0 +1,220 @@
+<!DOCTYPE html>
+<html>
+<head>
+    
+    <meta http-equiv="content-type" content="text/html; charset=UTF-8" />
+    <script src="https://cdn.jsdelivr.net/npm/leaflet@1.9.3/dist/leaflet.js"></script>
+    <script src="https://code.jquery.com/jquery-3.7.1.min.js"></script>
+    <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.2.2/dist/js/bootstrap.bundle.min.js"></script>
+    <script src="https://cdnjs.cloudflare.com/ajax/libs/Leaflet.awesome-markers/2.0.2/leaflet.awesome-markers.js"></script>
+    <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/leaflet@1.9.3/dist/leaflet.css"/>
+    <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/bootstrap@5.2.2/dist/css/bootstrap.min.css"/>
+    <link rel="stylesheet" href="https://netdna.bootstrapcdn.com/bootstrap/3.0.0/css/bootstrap-glyphicons.css"/>
+    <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/@fortawesome/fontawesome-free@6.2.0/css/all.min.css"/>
+    <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/Leaflet.awesome-markers/2.0.2/leaflet.awesome-markers.css"/>
+    <link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/python-visualization/folium/folium/templates/leaflet.awesome.rotate.min.css"/>
+    
+            <meta name="viewport" content="width=device-width,
+                initial-scale=1.0, maximum-scale=1.0, user-scalable=no" />
+            <style>
+                #map_78266fbe1759aea8d6c235f8af0c0d47 {
+                    position: relative;
+                    width: 100.0%;
+                    height: 100.0%;
+                    left: 0.0%;
+                    top: 0.0%;
+                }
+                .leaflet-container { font-size: 1rem; }
+            </style>
+
+            <style>html, body {
+                width: 100%;
+                height: 100%;
+                margin: 0;
+                padding: 0;
+            }
+            </style>
+
+            <style>#map {
+                position:absolute;
+                top:0;
+                bottom:0;
+                right:0;
+                left:0;
+                }
+            </style>
+
+            <script>
+                L_NO_TOUCH = false;
+                L_DISABLE_3D = false;
+            </script>
+
+        
+    <script src="https://cdnjs.cloudflare.com/ajax/libs/leaflet.draw/1.0.2/leaflet.draw.js"></script>
+    <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/leaflet.draw/1.0.2/leaflet.draw.css"/>
+</head>
+<body>
+    
+    
+            <div class="folium-map" id="map_78266fbe1759aea8d6c235f8af0c0d47" ></div>
+        
+    
+            
+            <style>
+                #export {
+                    position: absolute;
+                    top: 5px;
+                    right: 10px;
+                    z-index: 999;
+                    background: white;
+                    color: black;
+                    padding: 6px;
+                    border-radius: 4px;
+                    font-family: 'Helvetica Neue';
+                    cursor: pointer;
+                    font-size: 12px;
+                    text-decoration: none;
+                    top: 90px;
+                }
+            </style>
+            <a href='#' id='export'>Export</a>
+            
+        
+</body>
+<script>
+    
+    
+            var map_78266fbe1759aea8d6c235f8af0c0d47 = L.map(
+                "map_78266fbe1759aea8d6c235f8af0c0d47",
+                {
+                    center: [37.44975286338557, 126.65343455495042],
+                    crs: L.CRS.EPSG3857,
+                    ...{
+  "zoom": 17,
+  "zoomControl": true,
+  "preferCanvas": false,
+}
+
+                }
+            );
+
+            
+
+        
+    
+            var tile_layer_c1196f4c91b6a6393d3df89cf55b8c8f = L.tileLayer(
+                "https://tile.openstreetmap.org/{z}/{x}/{y}.png",
+                {
+  "minZoom": 0,
+  "maxZoom": 19,
+  "maxNativeZoom": 19,
+  "noWrap": false,
+  "attribution": "\u0026copy; \u003ca href=\"https://www.openstreetmap.org/copyright\"\u003eOpenStreetMap\u003c/a\u003e contributors",
+  "subdomains": "abc",
+  "detectRetina": false,
+  "tms": false,
+  "opacity": 1,
+}
+
+            );
+        
+    
+            tile_layer_c1196f4c91b6a6393d3df89cf55b8c8f.addTo(map_78266fbe1759aea8d6c235f8af0c0d47);
+        
+    
+        function geo_json_9b5f5972b80e71540402a2523e8a23d1_styler(feature) {
+            switch(feature.id) {
+                default:
+                    return {"color": "blue", "fill": false, "weight": 3};
+            }
+        }
+
+        function geo_json_9b5f5972b80e71540402a2523e8a23d1_onEachFeature(feature, layer) {
+
+            layer.on({
+            });
+        };
+        var geo_json_9b5f5972b80e71540402a2523e8a23d1 = L.geoJson(null, {
+                onEachFeature: geo_json_9b5f5972b80e71540402a2523e8a23d1_onEachFeature,
+            
+                style: geo_json_9b5f5972b80e71540402a2523e8a23d1_styler,
+            ...{
+}
+        });
+
+        function geo_json_9b5f5972b80e71540402a2523e8a23d1_add (data) {
+            geo_json_9b5f5972b80e71540402a2523e8a23d1
+                .addData(data);
+        }
+            geo_json_9b5f5972b80e71540402a2523e8a23d1_add({"features": [{"geometry": {"coordinates": [[[126.6504913, 37.4488536], [126.6502018, 37.4490381], [126.6501274, 37.4490101], [126.650006, 37.449035], [126.6499295, 37.4493095], [126.6497967, 37.4495766], [126.6496642, 37.4498464], [126.6493324, 37.4504076], [126.6490067, 37.4509487], [126.649781, 37.4514003], [126.6495914, 37.4517312], [126.6501326, 37.45175], [126.6505999, 37.4517726], [126.6510982, 37.4518266], [126.6512838, 37.4518467], [126.6517828, 37.4519342], [126.6520048, 37.4519731], [126.6520885, 37.4519878], [126.6524237, 37.4520465], [126.6525693, 37.452072], [126.6528855, 37.4521371], [126.6530797, 37.4521045], [126.6536373, 37.4518765], [126.6544021, 37.4516675], [126.6564271, 37.4511471], [126.6578465, 37.4507776], [126.6569217, 37.4492734], [126.6553859, 37.4466845], [126.6547144, 37.4468884], [126.6545294, 37.4469713], [126.6544155, 37.4470167], [126.653386, 37.4474272], [126.6527874, 37.4476634], [126.6507639, 37.4484888], [126.6505808, 37.4485626], [126.6505271, 37.4487565], [126.6504913, 37.4488536]]], "type": "Polygon"}, "id": "0", "type": "Feature"}], "type": "FeatureCollection"});
+
+        
+    
+            geo_json_9b5f5972b80e71540402a2523e8a23d1.addTo(map_78266fbe1759aea8d6c235f8af0c0d47);
+        
+    
+            var options = {
+              position: "topleft",
+              draw: {"circle": false, "circlemarker": false, "marker": false, "polygon": true, "polyline": false, "rectangle": true},
+              edit: {"edit": true, "remove": true},
+            }
+                // FeatureGroup is to store editable layers.
+                var drawnItems_draw_control_9d017ca223827f6de1a10b444cecfa37 =
+                    new L.featureGroup().addTo(
+                        map_78266fbe1759aea8d6c235f8af0c0d47
+                    );
+
+            options.edit.featureGroup = drawnItems_draw_control_9d017ca223827f6de1a10b444cecfa37;
+            var draw_control_9d017ca223827f6de1a10b444cecfa37 = new L.Control.Draw(
+                options
+            ).addTo( map_78266fbe1759aea8d6c235f8af0c0d47 );
+            map_78266fbe1759aea8d6c235f8af0c0d47.on(L.Draw.Event.CREATED, function(e) {
+                var layer = e.layer,
+                    type = e.layerType;
+                var coords = JSON.stringify(layer.toGeoJSON());
+                layer.on('click', function() {
+                    alert(coords);
+                    console.log(coords);
+                });
+                drawnItems_draw_control_9d017ca223827f6de1a10b444cecfa37.addLayer(layer);
+            });
+            map_78266fbe1759aea8d6c235f8af0c0d47.on('draw:created', function(e) {
+                drawnItems_draw_control_9d017ca223827f6de1a10b444cecfa37.addLayer(e.layer);
+            });
+
+            
+            document.getElementById('export').onclick = function(e) {
+                var data = drawnItems_draw_control_9d017ca223827f6de1a10b444cecfa37.toGeoJSON();
+                var convertedData = 'text/json;charset=utf-8,'
+                    + encodeURIComponent(JSON.stringify(data));
+                document.getElementById('export').setAttribute(
+                    'href', 'data:' + convertedData
+                );
+                document.getElementById('export').setAttribute(
+                    'download', "zones.geojson"
+                );
+            }
+            
+        
+    
+            var layer_control_859bc6ff55e56629719267d89d3e2a9c_layers = {
+                base_layers : {
+                    "openstreetmap" : tile_layer_c1196f4c91b6a6393d3df89cf55b8c8f,
+                },
+                overlays :  {
+                    "campus_boundary" : geo_json_9b5f5972b80e71540402a2523e8a23d1,
+                },
+            };
+            let layer_control_859bc6ff55e56629719267d89d3e2a9c = L.control.layers(
+                layer_control_859bc6ff55e56629719267d89d3e2a9c_layers.base_layers,
+                layer_control_859bc6ff55e56629719267d89d3e2a9c_layers.overlays,
+                {
+  "position": "topright",
+  "collapsed": true,
+  "autoZIndex": true,
+}
+            ).addTo(map_78266fbe1759aea8d6c235f8af0c0d47);
+
+        
+</script>
+</html>

pin.py

@@ -0,0 +1,376 @@
+import os
+import io
+
+import matplotlib
+matplotlib.use("Agg")  # 백엔드 Agg로 고정
+
+import geopandas as gpd
+import pandas as pd
+import numpy as np
+import matplotlib.pyplot as plt
+import contextily as ctx
+from shapely.geometry import Point
+from shapely import wkt as shp_wkt
+from PIL import Image
+
+import gradio as gr
+
+# -------------------- 경로 설정 --------------------
+SHP_PATH = r"C:\Users\USER\iCloudDrive\인하대학교\5학년 1학기\표본론 및 실습\TeamProject\Data\inha_boundary.shp"
+CSV_PATH = r"C:\Users\USER\iCloudDrive\인하대학교\5학년 1학기\표본론 및 실습\TeamProject\데이터 수집\측정결과.csv"
+
+# 이미지 해상도(픽셀)
+IMG_W, IMG_H = 1200, 900
+
+# -------------------- 캠퍼스 경계 준비 --------------------
+gdf_boundary_4326 = gpd.read_file(SHP_PATH).to_crs(epsg=4326)
+gdf_boundary_3857 = gdf_boundary_4326.to_crs(epsg=3857)
+
+# DeprecationWarning 피하기 위해 union_all 우선 사용
+try:
+    poly_3857 = gdf_boundary_3857.geometry.union_all()
+except AttributeError:
+    poly_3857 = gdf_boundary_3857.geometry.unary_union
+
+if poly_3857.geom_type == "MultiPolygon":
+    poly_3857 = max(poly_3857.geoms, key=lambda g: g.area)
+
+BOUND_MINX, BOUND_MINY, BOUND_MAXX, BOUND_MAXY = poly_3857.bounds
+
+# -------------------- 초기 CSV 로딩 --------------------
+# -------------------- 초기 CSV 로딩 --------------------
+if os.path.exists(CSV_PATH):
+    df_u = pd.read_csv(CSV_PATH, encoding="cp949")
+
+    # 컬럼 정리
+    df_u.columns = [c.strip() for c in df_u.columns]
+    df_u = df_u.loc[:, ~df_u.columns.str.contains("^Unnamed")]
+
+    if {"lat", "lon", "dB"}.issubset(df_u.columns):
+        # 이미 lat/lon/dB 형식이면 그대로 사용
+        df_init = df_u[["lat", "lon", "dB"]].copy()
+    elif "위치" in df_u.columns and "dB" in df_u.columns:
+        # 옛날 형식: 위치(WKT), dB → lat/lon/dB로 변환
+        pos_series = df_u["위치"].astype(str).str.strip()
+
+        def safe_load_wkt(s):
+            s_up = s.upper()
+            if "POINT" not in s_up:
+                return None
+            try:
+                return shp_wkt.loads(s)
+            except Exception:
+                return None
+
+        geom = pos_series.apply(safe_load_wkt)
+        mask = geom.notnull()
+
+        if mask.any():
+            geom_valid = geom[mask]
+            lons = geom_valid.apply(lambda g: round(g.x, 7))
+            lats = geom_valid.apply(lambda g: round(g.y, 7))
+            dB_vals = pd.to_numeric(df_u.loc[mask, "dB"], errors="coerce")
+
+            df_init = pd.DataFrame(
+                {
+                    "lat": lats.values,
+                    "lon": lons.values,
+                    "dB": dB_vals.values,
+                }
+            ).dropna(subset=["lat", "lon"])
+        else:
+            df_init = pd.DataFrame(columns=["lat", "lon", "dB"])
+    else:
+        # 형식을 알아볼 수 없으면 빈 DF
+        df_init = pd.DataFrame(columns=["lat", "lon", "dB"])
+else:
+    df_init = pd.DataFrame(columns=["lat", "lon", "dB"])
+
+# 타입 보정
+for col in ["lat", "lon", "dB"]:
+    if col not in df_init.columns:
+        df_init[col] = pd.Series(dtype=float)
+
+
+
+# -------------------- 지도 이미지 생성 --------------------
+def make_map_image(df_points: pd.DataFrame):
+    """캠퍼스 경계 + 저장된 포인트들을 그린 PNG 이미지를 numpy 배열로 반환."""
+    fig, ax = plt.subplots(
+        figsize=(IMG_W / 100, IMG_H / 100), dpi=100
+    )  # → 정확히 IMG_W x IMG_H 픽셀
+
+    # 캠퍼스 폴리곤 (테두리만)
+    gdf_boundary_3857.plot(
+        ax=ax,
+        facecolor="none",
+        edgecolor="darkblue",
+        linewidth=2,
+        alpha=1.0,
+    )
+
+    # 베이스맵
+    ctx.add_basemap(ax, source=ctx.providers.OpenStreetMap.Mapnik, alpha=0.7)
+
+    # 저장된 포인트 찍기
+    df_points = df_points.dropna(subset=["lat", "lon"])
+    if len(df_points) > 0:
+        gdf_pts = gpd.GeoDataFrame(
+            df_points.copy(),
+            geometry=gpd.points_from_xy(df_points["lon"], df_points["lat"]),
+            crs=4326,
+        ).to_crs(epsg=3857)
+
+        gdf_pts.plot(
+            ax=ax,
+            column="dB",
+            cmap="viridis",
+            markersize=40,
+            edgecolor="black",
+            linewidth=0.3,
+            alpha=0.9,
+            legend=False,
+            zorder=5,
+        )
+
+    # 폴리곤 bounds에 딱 맞게
+    ax.set_xlim(BOUND_MINX, BOUND_MAXX)
+    ax.set_ylim(BOUND_MINY, BOUND_MAXY)
+    ax.axis("off")
+
+    # 여백 제거해서 전체 캔버스를 지도에 맞춤
+    plt.subplots_adjust(0, 0, 1, 1)
+
+    # Figure -> PNG bytes -> PIL -> numpy
+    buf = io.BytesIO()
+    fig.savefig(buf, format="png", dpi=100)
+    buf.seek(0)
+    img = np.array(Image.open(buf))
+
+    plt.close(fig)
+    return img
+
+
+# -------------------- 픽셀 → 위경도 변환 --------------------
+def pixel_to_latlon(x_px: int, y_px: int):
+    """
+    Gradio 이미지 클릭 좌표(픽셀, (x, y))를
+    EPSG:3857 → EPSG:4326(lat, lon)으로 변환.
+    """
+    # x: [0, IMG_W] -> [BOUND_MINX, BOUND_MAXX]
+    x_3857 = BOUND_MINX + (BOUND_MAXX - BOUND_MINX) * (x_px / IMG_W)
+    # y: 위가 0, 아래가 IMG_H → [BOUND_MAXY, BOUND_MINY]로 매핑
+    y_3857 = BOUND_MAXY - (BOUND_MAXY - BOUND_MINY) * (y_px / IMG_H)
+
+    g = gpd.GeoSeries([Point(x_3857, y_3857)], crs=3857).to_crs(epsg=4326)
+    pt = g.iloc[0]
+    lat, lon = pt.y, pt.x
+
+    # 여기서 소수점 7자리로 반올림
+    return round(float(lat), 7), round(float(lon), 7)
+
+
+# -------------------- Gradio 콜백 --------------------
+def on_map_click(img, df_state, evt: gr.SelectData):
+    """이미지 클릭 시 위경도 텍스트 박스 갱신."""
+    if evt is None or evt.index is None:
+        return None, None
+    x, y = evt.index  # (x_px, y_px)
+    lat, lon = pixel_to_latlon(x, y)
+    return lat, lon  # 이미 7자리로 반올림된 값
+
+
+def add_point(lat, lon, db, df_state):
+    """lat/lon/dB 추가하고 CSV 저장 + 지도 갱신."""
+    if df_state is None or isinstance(df_state, dict):
+        df_state = df_init.copy()
+
+    if lat is None or lon is None or db is None:
+        msg = "위도/경도/dB 입력 오류."
+        return df_state, make_map_image(df_state), df_state, msg
+
+    try:
+        lat = float(lat)
+        lon = float(lon)
+        db = float(db)
+    except ValueError:
+        msg = "숫자 오류."
+        return df_state, make_map_image(df_state), df_state, msg
+
+    # 안전하게 한 번 더 7자리로 맞춤
+    lat = round(lat, 7)
+    lon = round(lon, 7)
+
+    # 폴리곤 안인지 체크
+    pt = gpd.GeoSeries([Point(lon, lat)], crs=4326).to_crs(epsg=3857).iloc[0]
+    if not poly_3857.contains(pt):
+        msg = "캠퍼스 경계 밖."
+        return df_state, make_map_image(df_state), df_state, msg
+
+    new_row = pd.DataFrame([{"lat": lat, "lon": lon, "dB": db}])
+    new_df = pd.concat([df_state, new_row], ignore_index=True)
+
+    os.makedirs(os.path.dirname(CSV_PATH), exist_ok=True)
+    new_df[["lat", "lon", "dB"]].to_csv(CSV_PATH, index=False, encoding="cp949")
+
+    # 메시지도 7자리로
+    msg = f"{len(new_df)}개 점 저장됨 (마지막: lat={lat:.7f}, lon={lon:.7f}, dB={db:.1f})"
+    return new_df, make_map_image(new_df), new_df, msg
+
+def load_uploaded_csv(file, df_state):
+    """
+    CSV 업로드 시:
+    - (케이스 A) '위치', 'dB' 컬럼이 있으면: 위치(WKT) -> lat/lon, dB 그대로
+    - (케이스 B) 'lat', 'lon', 'dB' 컬럼이 있으면: 그대로 사용
+    - 둘 다 아니면 에러
+    """
+    if df_state is None or isinstance(df_state, dict):
+        df_state = df_init.copy()
+
+    if file is None:
+        msg = "업로드된 파일 없음."
+        return df_state, make_map_image(df_state), df_state, msg
+
+    # CSV 읽기 (인코딩 시도)
+    try:
+        df_u = pd.read_csv(file.name, encoding="cp949")
+    except UnicodeDecodeError:
+        df_u = pd.read_csv(file.name, encoding="utf-8")
+
+    # 컬럼 이름 공백/Unnamed 정리
+    df_u.columns = [c.strip() for c in df_u.columns]
+    df_u = df_u.loc[:, ~df_u.columns.str.contains("^Unnamed")]
+
+    # ---------- 케이스 B: lat/lon/dB 형식 ----------
+    if {"lat", "lon", "dB"}.issubset(df_u.columns):
+        new_df = df_u[["lat", "lon", "dB"]].copy()
+        # 숫자/반올림 보정
+        new_df["lat"] = pd.to_numeric(new_df["lat"], errors="coerce").round(7)
+        new_df["lon"] = pd.to_numeric(new_df["lon"], errors="coerce").round(7)
+        new_df["dB"] = pd.to_numeric(new_df["dB"], errors="coerce")
+        new_df = new_df.dropna(subset=["lat", "lon"])
+
+        df_merged = new_df.copy()
+
+        os.makedirs(os.path.dirname(CSV_PATH), exist_ok=True)
+        df_merged[["lat", "lon", "dB"]].to_csv(CSV_PATH, index=False, encoding="cp949")
+
+        msg = f"lat/lon/dB 형식 CSV에서 {len(df_merged)}개 점 읽어옴."
+        return df_merged, make_map_image(df_merged), df_merged, msg
+
+    # ---------- 케이스 A: 위치(WKT), dB 형식 ----------
+    if "위치" in df_u.columns and "dB" in df_u.columns:
+        pos_series = df_u["위치"].astype(str).str.strip()
+
+        def safe_load_wkt(s):
+            s_up = s.upper()
+            if "POINT" not in s_up:
+                return None
+            try:
+                return shp_wkt.loads(s)
+            except Exception:
+                return None
+
+        geom = pos_series.apply(safe_load_wkt)
+        mask = geom.notnull()
+
+        if not mask.any():
+            msg = "'위치' 컬럼에 유효한 POINT WKT가 없다."
+            return df_state, make_map_image(df_state), df_state, msg
+
+        geom_valid = geom[mask]
+        lons = geom_valid.apply(lambda g: round(g.x, 7))
+        lats = geom_valid.apply(lambda g: round(g.y, 7))
+        dB_vals = pd.to_numeric(df_u.loc[mask, "dB"], errors="coerce")
+
+        new_df = pd.DataFrame(
+            {
+                "lat": lats.values,
+                "lon": lons.values,
+                "dB": dB_vals.values,
+            }
+        ).dropna(subset=["lat", "lon"])
+
+        df_merged = new_df.copy()
+
+        os.makedirs(os.path.dirname(CSV_PATH), exist_ok=True)
+        df_merged[["lat", "lon", "dB"]].to_csv(CSV_PATH, index=False, encoding="cp949")
+
+        msg = f"[위치, dB] 형식 CSV에서 유효한 점 {len(df_merged)}개 읽어옴."
+        return df_merged, make_map_image(df_merged), df_merged, msg
+
+    # ---------- 둘 다 아니면 ----------
+    msg = "CSV 형식을 알아볼 수 없다. (위치,dB 또는 lat,lon,dB 필요)"
+    return df_state, make_map_image(df_state), df_state, msg
+
+
+
+# -------------------- Gradio 앱 --------------------
+def build_app():
+    init_img = make_map_image(df_init)
+
+    with gr.Blocks(title="소음 측정 기록기") as demo:
+        gr.Markdown(
+            "## 인하대 캠퍼스 소음 측정\n"
+            "- 지도 이미지 클릭 → 위/경도 자동 입력 → dB 저장\n"
+            "- 또는 [위치, dB] CSV 업로드해서 한 번에 보여주기"
+        )
+
+        df_state = gr.State(df_init)
+
+        with gr.Row():
+            map_img = gr.Image(
+                value=init_img,
+                type="numpy",
+                label=None,
+                interactive=True,
+            )
+
+            with gr.Column():
+                lat_box = gr.Number(label="위도 (lat)", interactive=False)
+                lon_box = gr.Number(label="경도 (lon)", interactive=False)
+                db_box = gr.Number(label="데시벨 (dB)")
+
+                add_btn = gr.Button("현재 좌표 추가 + CSV 저장", variant="primary")
+                csv_upload = gr.File(
+                    label="CSV 업로드 ([위치, dB] 형식)",
+                    file_types=[".csv"],
+                )
+
+                status = gr.Markdown()
+
+                table = gr.Dataframe(
+                    value=df_init,
+                    headers=["lat", "lon", "dB"],
+                    label="저장된 측정값",
+                    interactive=False,
+                    wrap=True,
+                )
+
+        # 이미지 클릭 시: lat/lon 갱신
+        map_img.select(
+            fn=on_map_click,
+            inputs=[map_img, df_state],
+            outputs=[lat_box, lon_box],
+        )
+
+        # 버튼 클릭 시: DF/지도/상태 갱신 + CSV 저장
+        add_btn.click(
+            fn=add_point,
+            inputs=[lat_box, lon_box, db_box, df_state],
+            outputs=[table, map_img, df_state, status],
+        )
+
+        # CSV 업로드 시: 업로드 데이터로 지도/테이블/상태 갱신
+        csv_upload.change(
+            fn=load_uploaded_csv,
+            inputs=[csv_upload, df_state],
+            outputs=[table, map_img, df_state, status],
+        )
+
+    return demo
+
+
+if __name__ == "__main__":
+    app = build_app()
+    app.launch(share=True)

simul.py

@@ -0,0 +1,539 @@
+# sampling_sim_app.py
+import os
+import io
+
+import matplotlib
+matplotlib.use("Agg")
+
+import numpy as np
+import pandas as pd
+import geopandas as gpd
+import matplotlib.pyplot as plt
+from shapely.geometry import Point
+from shapely import wkt as shp_wkt
+from PIL import Image
+import contextily as ctx  # ← 추가
+
+import gradio as gr
+
+# ----------------- 경로 설정 -----------------
+BASE_DIR = os.path.dirname(__file__)
+DATA_DIR = os.path.join(BASE_DIR, "Data")
+
+SHP_PATH = os.path.join(DATA_DIR, "inha_boundary.shp")
+ZONES_PATH = os.path.join(DATA_DIR, "zones_최종.geojson")
+
+# 측정결과 CSV (lat, lon, dB 또는 위치, dB)
+CSV_PATH = os.path.join(DATA_DIR, "측정결과_최종.csv")
+
+
+# ---------- 유틸: Figure -> numpy ----------
+def fig_to_array(fig, dpi=100):
+    buf = io.BytesIO()
+    fig.savefig(buf, format="png", dpi=dpi, bbox_inches="tight")
+    buf.seek(0)
+    img = np.array(Image.open(buf))
+    plt.close(fig)
+    return img
+
+
+# ---------- 경계 SHP 로딩 ----------
+def load_boundary():
+    gdf4326 = gpd.read_file(SHP_PATH).to_crs(epsg=4326)
+    gdf3857 = gdf4326.to_crs(epsg=3857)
+
+    poly3857 = gdf3857.geometry.unary_union
+    if poly3857.geom_type == "MultiPolygon":
+        poly3857 = max(poly3857.geoms, key=lambda g: g.area)
+
+    return gdf4326, gdf3857, poly3857
+
+
+# ---------- 측정결과 CSV 로딩 ----------
+def load_points_from_csv():
+    try:
+        df_u = pd.read_csv(CSV_PATH, encoding="cp949")
+    except UnicodeDecodeError:
+        df_u = pd.read_csv(CSV_PATH, encoding="utf-8")
+
+    df_u.columns = [c.strip() for c in df_u.columns]
+    df_u = df_u.loc[:, ~df_u.columns.str.contains("^Unnamed")]
+
+    # 케이스 B: 위치, dB
+    if "위치" in df_u.columns and "dB" in df_u.columns:
+        pos_series = df_u["위치"].astype(str).str.strip()
+
+        def safe_load_wkt(s):
+            s_up = s.upper()
+            if "POINT" not in s_up:
+                return None
+            try:
+                return shp_wkt.loads(s)
+            except Exception:
+                return None
+
+        geom = pos_series.apply(safe_load_wkt)
+        mask = geom.notnull()
+        if not mask.any():
+            raise ValueError("'위치' 컬럼에 유효한 POINT WKT가 없다.")
+
+        geom_valid = geom[mask]
+        lons = geom_valid.apply(lambda g: round(g.x, 7))
+        lats = geom_valid.apply(lambda g: round(g.y, 7))
+        dBs = pd.to_numeric(df_u.loc[mask, "dB"], errors="coerce")
+
+        df = pd.DataFrame({"lat": lats, "lon": lons, "dB": dBs}).dropna(
+            subset=["lat", "lon"]
+        )
+        return df
+
+    # 케이스 A: lat, lon, dB
+    if {"lat", "lon", "dB"}.issubset(df_u.columns):
+        df = df_u[["lat", "lon", "dB"]].copy()
+        df["lat"] = pd.to_numeric(df["lat"], errors="coerce").round(7)
+        df["lon"] = pd.to_numeric(df["lon"], errors="coerce").round(7)
+        df["dB"] = pd.to_numeric(df["dB"], errors="coerce")
+        df = df.dropna(subset=["lat", "lon"])
+        return df
+
+    raise ValueError("CSV는 (lat, lon, dB) 또는 (위치, dB) 형식이어야 한다.")
+
+
+# ---------- 구역 GeoJSON 로딩 ----------
+def load_zones(poly3857):
+    zones4326 = gpd.read_file(ZONES_PATH).to_crs(epsg=4326)
+    zones3857 = zones4326.to_crs(epsg=3857)
+
+    if "zone_id" not in zones3857.columns:
+        zones3857["zone_id"] = np.arange(1, len(zones3857) + 1)
+
+    zones3857["geom_clip"] = zones3857.geometry.intersection(poly3857)
+    zones3857["area"] = zones3857["geom_clip"].area
+    zones3857 = zones3857[zones3857["area"] > 0].copy()
+
+    A_total = zones3857["area"].sum()
+    zones3857["W_area"] = zones3857["area"] / A_total
+
+    return zones4326, zones3857
+
+
+# ---------- 포인트에 zone_id 붙이기 ----------
+def attach_zone_id(points_df, zones3857):
+    pts4326 = gpd.GeoDataFrame(
+        points_df.copy(),
+        geometry=gpd.points_from_xy(points_df["lon"], points_df["lat"]),
+        crs=4326,
+    )
+    pts3857 = pts4326.to_crs(epsg=3857)
+
+    zones_clip = zones3857[["zone_id", "geom_clip"]].copy()
+    zones_clip = zones_clip.set_geometry("geom_clip")
+    zones_clip = zones_clip.rename_geometry("geometry")
+
+    joined = gpd.sjoin(
+        pts3857,
+        zones_clip[["zone_id", "geometry"]],
+        how="inner",
+        predicate="within",
+    )
+
+    out = joined[["lat", "lon", "dB", "zone_id"]].copy().reset_index(drop=True)
+    return out
+
+
+# ---------- 공통 준비: 경계/구역/포인트/층정보 ----------
+def prepare_sampling_data():
+    boundary4326, boundary3857, poly3857 = load_boundary()
+    zones4326, zones3857 = load_zones(poly3857)
+    df_points = load_points_from_csv()
+    points_zone = attach_zone_id(df_points, zones3857)
+
+    # 층별 모집단 크기 N_h
+    zone_sizes = points_zone.groupby("zone_id").size().rename("N_h")
+    zones_info = zones3857[["zone_id", "area", "W_area"]].merge(
+        zone_sizes, on="zone_id", how="left"
+    )
+    zones_info["N_h"] = zones_info["N_h"].fillna(0).astype(int)
+
+    # 층별 표준편차 S_h (네이만 배분용)
+    strata_std = points_zone.groupby("zone_id")["dB"].std(ddof=1).rename("S_h")
+    zones_info = zones_info.merge(strata_std, on="zone_id", how="left")
+    zones_info["S_h"] = zones_info["S_h"].fillna(0.0)
+
+    return boundary4326, zones4326, df_points, points_zone, zones_info
+
+
+# ---------- 면적비례 n_h 계산 ----------
+def allocate_n_by_area(zones_info, n_total):
+    zones = zones_info.copy()
+    zones["n_raw"] = n_total * zones["W_area"]
+    zones["n"] = zones["n_raw"].round().astype(int)
+
+    diff = n_total - zones["n"].sum()
+    if diff != 0:
+        order = zones["n_raw"].sort_values(ascending=(diff < 0)).index
+        for idx in order[:abs(diff)]:
+            zones.loc[idx, "n"] += 1 if diff > 0 else -1
+
+    zones["n"] = zones[["n", "N_h"]].min(axis=1)
+    return zones
+
+
+# ---------- 네이만 배분 n_h 계산 ----------
+def allocate_n_neyman(zones_info, n_total):
+    zones = zones_info.copy()
+    if "S_h" not in zones.columns:
+        raise ValueError("zones_info에 S_h가 없다. 네이만 배분 전에 S_h를 계산해야 한다.")
+
+    zones["S_h"] = zones["S_h"].fillna(0.0)
+
+    # 가중치 ∝ N_h * S_h
+    weight = zones["N_h"] * zones["S_h"]
+    if weight.sum() <= 0:
+        # 분산이 전부 0이면 N_h 비례로 배분
+        weight = zones["N_h"].copy()
+
+    zones["n_raw"] = n_total * weight / weight.sum()
+    zones["n"] = zones["n_raw"].round().astype(int)
+
+    diff = n_total - zones["n"].sum()
+    if diff != 0:
+        order = zones["n_raw"].sort_values(ascending=(diff < 0)).index
+        for idx in order[:abs(diff)]:
+            zones.loc[idx, "n"] += 1 if diff > 0 else -1
+
+    zones["n"] = zones[["n", "N_h"]].min(axis=1)
+    return zones
+
+
+# ---------- 시뮬레이션 ----------
+def simulate_sampling(points_zone_df, zones_info, n, R, strat_label):
+    y_all = points_zone_df["dB"].to_numpy()
+    N = len(y_all)
+    mu_pop = float(np.mean(y_all))
+
+    zones = zones_info.set_index("zone_id")
+    W_h = zones["N_h"] / zones["N_h"].sum()
+    n_h = zones["n"]
+
+    idx_by_zone = {}
+    for z in zones.index:
+        idx_by_zone[z] = points_zone_df.index[
+            points_zone_df["zone_id"] == z
+        ].to_numpy()
+
+    srs_means = []
+    strat_means = []
+
+    for _ in range(R):
+        # SRS
+        n_eff = min(n, N)
+        idx_srs = np.random.choice(N, size=n_eff, replace=False)
+        mu_srs = float(np.mean(y_all[idx_srs]))
+        srs_means.append(mu_srs)
+
+        # Stratified
+        mu_str = 0.0
+        for z in zones.index:
+            N_h = int(zones.loc[z, "N_h"])
+            nh = int(n_h.loc[z])
+            if N_h == 0 or nh == 0:
+                continue
+            idx_pool = idx_by_zone[z]
+            if len(idx_pool) < nh:
+                nh = len(idx_pool)
+            choose = np.random.choice(idx_pool, size=nh, replace=False)
+            y_h = points_zone_df.loc[choose, "dB"].to_numpy()
+            ybar_h = float(np.mean(y_h))
+            mu_str += float(W_h.loc[z]) * ybar_h
+        strat_means.append(mu_str)
+
+    srs_means = np.array(srs_means)
+    strat_means = np.array(strat_means)
+
+    def metrics(arr, name):
+        mean_hat = float(arr.mean())
+        var_hat = float(arr.var(ddof=1))
+        bias = mean_hat - mu_pop
+        mse = float(((arr - mu_pop) ** 2).mean())
+        return {
+            "design": name,
+            "mean_hat": mean_hat,
+            "var_hat": var_hat,
+            "bias": bias,
+            "MSE": mse,
+        }
+
+    m_srs = metrics(srs_means, "SRS")
+    m_str = metrics(strat_means, strat_label)
+
+    df_metrics = pd.DataFrame([m_srs, m_str])
+    df_metrics["deff_vs_SRS"] = df_metrics["var_hat"] / df_metrics["var_hat"].iloc[0]
+
+    # 숫자 4째 자리까지 반올림
+    num_cols = ["mean_hat", "var_hat", "bias", "MSE", "deff_vs_SRS"]
+    df_metrics[num_cols] = df_metrics[num_cols].round(4)
+
+    return mu_pop, srs_means, strat_means, df_metrics
+
+
+# ---------- 구역 요약 테이블 ----------
+def build_zones_table(zones_alloc):
+    zones_table = zones_alloc[["zone_id", "area", "W_area", "N_h", "n"]].copy()
+    zones_table = zones_table.rename(
+        columns={
+            "area": "area_m2",
+            "W_area": "area_weight",
+            "N_h": "N_pop",
+            "n": "n_alloc",
+        }
+    )
+    zones_table["sampling_frac"] = zones_table["n_alloc"] / zones_table["N_pop"]
+
+    total_area_m2 = zones_table["area_m2"].sum()
+    total_area_weight = zones_table["area_weight"].sum()
+    total_N = zones_table["N_pop"].sum()
+    total_n = zones_table["n_alloc"].sum()
+    total_sampling_frac = total_n / total_N if total_N > 0 else np.nan
+
+    total_row = {
+        "zone_id": "���계",
+        "area_m2": total_area_m2,
+        "area_weight": total_area_weight,
+        "N_pop": total_N,
+        "n_alloc": total_n,
+        "sampling_frac": total_sampling_frac,
+    }
+
+    zones_table = zones_table[
+        ["zone_id", "area_m2", "area_weight", "N_pop", "n_alloc", "sampling_frac"]
+    ]
+    zones_table = pd.concat([zones_table, pd.DataFrame([total_row])], ignore_index=True)
+
+    # 숫자 컬럼 4째 자리까지 반올림
+    num_cols_zone = ["area_m2", "area_weight", "N_pop", "n_alloc", "sampling_frac"]
+    for col in num_cols_zone:
+        zones_table[col] = pd.to_numeric(zones_table[col], errors="coerce").round(4)
+
+    return zones_table, float(total_N), float(total_n)
+
+
+# ---------- 시각화 ----------
+def plot_population_map(boundary4326, zones4326, points_df):
+    g_boundary = boundary4326.to_crs(epsg=3857)
+    g_zones = zones4326.to_crs(epsg=3857)
+    g_pts = gpd.GeoDataFrame(
+        points_df.copy(),
+        geometry=gpd.points_from_xy(points_df["lon"], points_df["lat"]),
+        crs=4326,
+    ).to_crs(epsg=3857)
+
+    fig, ax = plt.subplots(figsize=(8, 8))
+
+    # 베이스맵 먼저 깔기
+    minx, miny, maxx, maxy = g_boundary.total_bounds
+    ax.set_xlim(minx, maxx)
+    ax.set_ylim(miny, maxy)
+    ctx.add_basemap(ax, source=ctx.providers.OpenStreetMap.Mapnik, alpha=0.7)
+
+    # 그 위에 경계/구역/포인트
+    g_boundary.plot(ax=ax, facecolor="none", edgecolor="black", linewidth=2)
+    if len(g_zones) > 0:
+        g_zones.plot(
+            ax=ax,
+            facecolor="none",
+            edgecolor="red",
+            linewidth=1.5,
+            alpha=0.7,
+        )
+
+    g_pts.plot(
+        ax=ax,
+        column="dB",
+        cmap="viridis",
+        markersize=20,
+        edgecolor="black",
+        linewidth=0.2,
+        legend=False,
+        alpha=0.9,
+    )
+
+    ax.set_title("Population points with zones", fontsize=12)
+    ax.axis("off")
+    plt.tight_layout()
+    return fig_to_array(fig)
+
+
+def plot_histograms(srs_means, strat_means, mu_pop, strat_label):
+    fig, ax = plt.subplots(figsize=(8, 5))
+
+    ax.hist(srs_means, bins=20, alpha=0.5, label="SRS")
+    ax.hist(strat_means, bins=20, alpha=0.5, label=strat_label)
+    ax.axvline(mu_pop, color="k", linestyle="--", label="Population mean")
+
+    ax.set_xlabel("Sample mean (dB)")
+    ax.set_ylabel("Frequency")
+    ax.legend()
+    ax.set_title("Sampling distribution of mean")
+    plt.tight_layout()
+    return fig_to_array(fig)
+
+
+# ---------- Gradio 콜백: 면적비례 ----------
+def run_sim_area(n, R):
+    n = int(n)
+    R = int(R)
+
+    boundary4326, zones4326, df_points, points_zone, zones_info = prepare_sampling_data()
+
+    # 면적비례 배분
+    zones_alloc = allocate_n_by_area(zones_info, n)
+    zones_table, total_N, total_n = build_zones_table(zones_alloc)
+
+    mu_pop, srs_means, strat_means, df_metrics = simulate_sampling(
+        points_zone, zones_alloc, n, R, "Stratified (area-proportional)"
+    )
+
+    map_img = plot_population_map(boundary4326, zones4326, df_points)
+    hist_img = plot_histograms(
+        srs_means, strat_means, mu_pop, "Stratified (area-proportional)"
+    )
+
+    summary = (
+        f"[면적비례 층화]\n"
+        f"모집단 포인트 수: {int(total_N)}개\n"
+        f"구역 수: {len(zones_alloc)}개\n"
+        f"모집단 평균 dB: {mu_pop:.4f}\n"
+        f"총 표본크기 n = {int(total_n)} (입력값 {n}), 반복 R = {R}"
+    )
+
+    return map_img, hist_img, df_metrics, zones_table, summary
+
+
+# ---------- Gradio 콜백: 네이만 ----------
+def run_sim_neyman(n, R):
+    n = int(n)
+    R = int(R)
+
+    boundary4326, zones4326, df_points, points_zone, zones_info = prepare_sampling_data()
+
+    # 네이만 배분
+    zones_alloc = allocate_n_neyman(zones_info, n)
+    zones_table, total_N, total_n = build_zones_table(zones_alloc)
+
+    mu_pop, srs_means, strat_means, df_metrics = simulate_sampling(
+        points_zone, zones_alloc, n, R, "Stratified (Neyman)"
+    )
+
+    map_img = plot_population_map(boundary4326, zones4326, df_points)
+    hist_img = plot_histograms(
+        srs_means, strat_means, mu_pop, "Stratified (Neyman)"
+    )
+
+    summary = (
+        f"[네이만 층화]\n"
+        f"모집단 포인트 수: {int(total_N)}개\n"
+        f"구역 수: {len(zones_alloc)}개\n"
+        f"모집단 평균 dB: {mu_pop:.4f}\n"
+        f"총 표본크기 n = {int(total_n)} (입력값 {n}), 반복 R = {R}"
+    )
+
+    return map_img, hist_img, df_metrics, zones_table, summary
+
+
+# ---------- Gradio 앱 ----------
+def build_app():
+    with gr.Blocks(title="표본추출 시뮬레이션 (SRS vs 층화)") as demo:
+
+        with gr.Tab("SRS vs 면적비례 층화"):
+            with gr.Row():
+                n_area = gr.Number(label="표본크기 n", value=60, precision=0)
+                R_area = gr.Number(label="시뮬레이션 반복 횟수 R", value=500, precision=0)
+
+            run_btn_area = gr.Button("면적비례 층화 시뮬레이션 실행", variant="primary")
+
+            with gr.Row():
+                map_img_area = gr.Image(label="Population + zones")
+                hist_img_area = gr.Image(label="Sampling distribution")
+
+            metrics_table_area = gr.Dataframe(
+                headers=[
+                    "design",
+                    "mean_hat",
+                    "var_hat",
+                    "bias",
+                    "MSE",
+                    "deff_vs_SRS",
+                ],
+                label="성능 비교 지표",
+                interactive=False,
+            )
+
+            zones_table_area = gr.Dataframe(
+                label="구역별 면적 및 모집단/배정 표본수",
+                interactive=False,
+            )
+
+            summary_md_area = gr.Markdown()
+
+            run_btn_area.click(
+                fn=run_sim_area,
+                inputs=[n_area, R_area],
+                outputs=[
+                    map_img_area,
+                    hist_img_area,
+                    metrics_table_area,
+                    zones_table_area,
+                    summary_md_area,
+                ],
+            )
+
+        with gr.Tab("SRS vs 네이만 층화"):
+            with gr.Row():
+                n_neyman = gr.Number(label="표본크기 n", value=60, precision=0)
+                R_neyman = gr.Number(label="시뮬레이션 반복 횟수 R", value=500, precision=0)
+
+            run_btn_neyman = gr.Button("네이만 층화 시뮬레이션 실행", variant="primary")
+
+            with gr.Row():
+                map_img_neyman = gr.Image(label="Population + zones")
+                hist_img_neyman = gr.Image(label="Sampling distribution")
+
+            metrics_table_neyman = gr.Dataframe(
+                headers=[
+                    "design",
+                    "mean_hat",
+                    "var_hat",
+                    "bias",
+                    "MSE",
+                    "deff_vs_SRS",
+                ],
+                label="성능 비교 지표",
+                interactive=False,
+            )
+
+            zones_table_neyman = gr.Dataframe(
+                label="구역별 면적 및 모집단/배정 표본수",
+                interactive=False,
+            )
+
+            summary_md_neyman = gr.Markdown()
+
+            run_btn_neyman.click(
+                fn=run_sim_neyman,
+                inputs=[n_neyman, R_neyman],
+                outputs=[
+                    map_img_neyman,
+                    hist_img_neyman,
+                    metrics_table_neyman,
+                    zones_table_neyman,
+                    summary_md_neyman,
+                ],
+            )
+
+    return demo
+
+
+if __name__ == "__main__":
+    app = build_app()
+    app.launch(share=True)

tset.ipynb

@@ -0,0 +1,1135 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "9afdef8a",
+   "metadata": {},
+   "source": [
+    "# 경계정보 생성"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "id": "014473d8",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "🔍 인하대학교 경계 정보 추출 중...\n",
+      "응답 상태 코드: 504\n",
+      "❌ API 오류 발생\n",
+      "🔄 대체 방법을 시도합니다...\n",
+      "📍 Plan B: 좌표 범위 검색 방법...\n",
+      "✓ 대체 방법 성공!\n",
+      "✓ 3개의 요소 발견\n",
+      "🔍 받은 요소 개수: 3\n",
+      "  [0] type: way, tags: {'amenity': 'university', 'building': 'yes', 'education': 'university', 'name': '나빌레관', 'name:en': 'Student Club Hall'}\n",
+      "      좌표 개수: 13\n",
+      "      면적: 0.00000008\n",
+      "      ✓ 추가됨\n",
+      "  [1] type: way, tags: {'amenity': 'university', 'education': 'university', 'name': '인하대학교', 'name:en': 'Inha University', 'name:ja': '仁荷大学校', 'name:zh': '仁荷大学', 'wikidata': 'Q488851'}\n",
+      "      좌표 개수: 37\n",
+      "      면적: 0.00002906\n",
+      "      ✓ 추가됨\n",
+      "  [2] type: way, tags: {'amenity': 'university', 'education': 'university', 'name': '인하공업전문대학', 'name:en': 'Inha Technical College', 'name:ja': '仁荷工業専門大学', 'name:zh': '仁荷工业专门大学'}\n",
+      "      좌표 개수: 26\n",
+      "      면적: 0.00001058\n",
+      "      ✓ 추가됨\n",
+      "📊 총 3개 feature 생성\n",
+      "   가장 큰 polygon 선택 (면적: 0.00002906)\n",
+      "✓ GeoJSON 파일 저장 완료: inha_boundary.geojson\n",
+      "✓ 위성 이미지 (Esri) 추가 완료\n",
+      "✓ 위성 이미지 저장 완료: inha_boundary_satellite.png\n",
+      "✓ 지도 스타일 + 클리핑 추가 완료\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Users\\USER\\AppData\\Local\\Temp\\ipykernel_4744\\1166856351.py:227: DeprecationWarning: The 'unary_union' attribute is deprecated, use the 'union_all()' method instead.\n",
+      "  poly = gdf_web.geometry.unary_union\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "✓ 지도 이미지 저장 완료: inha_boundary_map.png\n",
+      "✓ 폴리곤 이미지 저장 완료: inha_boundary_polygon.png\n",
+      "✓ Shapefile 저장 완료: inha_boundary.shp\n",
+      "\n",
+      "📍 경계 정보:\n",
+      "   - 서쪽: 126.649007, 남쪽: 37.446685\n",
+      "   - 동쪽: 126.657847, 북쪽: 37.452137\n",
+      "   - 중심점: (126.653435, 37.449753)\n",
+      "\n",
+      "📌 첫 번째 경계의 좌표:\n",
+      "   총 37개의 좌표점\n",
+      "   시작점: (126.6504913, 37.4488536)\n",
+      "   끝점: (126.6504913, 37.4488536)\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "c:\\Users\\USER\\anaconda3\\envs\\torch_ev\\lib\\site-packages\\pyogrio\\raw.py:723: RuntimeWarning: Normalized/laundered field name: 'name:en' to 'name_en'\n",
+      "  ogr_write(\n",
+      "c:\\Users\\USER\\anaconda3\\envs\\torch_ev\\lib\\site-packages\\pyogrio\\raw.py:723: RuntimeWarning: Normalized/laundered field name: 'name:ja' to 'name_ja'\n",
+      "  ogr_write(\n",
+      "c:\\Users\\USER\\anaconda3\\envs\\torch_ev\\lib\\site-packages\\pyogrio\\raw.py:723: RuntimeWarning: Normalized/laundered field name: 'name:zh' to 'name_zh'\n",
+      "  ogr_write(\n"
+     ]
+    }
+   ],
+   "source": [
+    "import numpy as np\n",
+    "import matplotlib.path as mpath\n",
+    "import matplotlib.patches as mpatches\n",
+    "import requests\n",
+    "import json\n",
+    "import geopandas as gpd\n",
+    "from shapely.geometry import shape, Polygon, MultiPolygon\n",
+    "import matplotlib.pyplot as plt\n",
+    "import contextily as ctx\n",
+    "from matplotlib import font_manager\n",
+    "\n",
+    "# 대체 방법: 좌표 범위로 검색\n",
+    "def get_inha_boundary_alternative():\n",
+    "    print(\"📍 Plan B: 좌표 범위 검색 방법...\")\n",
+    "    overpass_url = \"http://overpass-api.de/api/interpreter\"\n",
+    "    \n",
+    "    # 인하대학교 주변 영역 (위도, 경도 범위)\n",
+    "    overpass_query = f\"\"\"\n",
+    "    [out:json][timeout:25];\n",
+    "    (\n",
+    "      way[\"amenity\"=\"university\"](37.445,126.650,37.456,126.662);\n",
+    "      relation[\"amenity\"=\"university\"](37.445,126.650,37.456,126.662);\n",
+    "      way[\"name~\"=\"인하\"](37.445,126.650,37.456,126.662);\n",
+    "    );\n",
+    "    out geom;\n",
+    "    \"\"\"\n",
+    "    \n",
+    "    try:\n",
+    "        response = requests.post(overpass_url, data={'data': overpass_query}, timeout=30)\n",
+    "        if response.status_code == 200:\n",
+    "            data = response.json()\n",
+    "            if data.get('elements'):\n",
+    "                print(\"✓ 대체 방법 성공!\")\n",
+    "                return data\n",
+    "    except Exception as e:\n",
+    "        print(f\"⚠️ 대체 방법 실패: {e}\")\n",
+    "    \n",
+    "    print(\"📍 Plan C: 수동 경계 생성...\")\n",
+    "    return create_manual_boundary()\n",
+    "\n",
+    "# 수동으로 경계 생성 (실제 인하대 메인 캠퍼스만)\n",
+    "def create_manual_boundary():\n",
+    "    print(\"✅ 인하대학교 메인 캠퍼스 경계 생성\")\n",
+    "    # 인하대학교 메인 캠퍼스 경계 (정문-5호관-60주년기념관-본관-체육관-후문 순)\n",
+    "    coords = [\n",
+    "        [126.65168, 37.45100],  # 정문\n",
+    "        [126.65150, 37.45095],\n",
+    "        [126.65140, 37.45020],\n",
+    "        [126.65100, 37.44920],\n",
+    "        [126.65090, 37.44880],\n",
+    "        [126.65120, 37.44860],\n",
+    "        [126.65280, 37.44780],\n",
+    "        [126.65450, 37.44680],\n",
+    "        [126.65580, 37.44680],  # 남동쪽 모서리\n",
+    "        [126.65750, 37.44720],\n",
+    "        [126.65850, 37.44780],\n",
+    "        [126.65900, 37.44860],\n",
+    "        [126.65920, 37.44980],\n",
+    "        [126.65900, 37.45100],  # 동쪽 경계\n",
+    "        [126.65850, 37.45180],\n",
+    "        [126.65750, 37.45240],\n",
+    "        [126.65620, 37.45260],  # 북동쪽 모서리\n",
+    "        [126.65500, 37.45250],\n",
+    "        [126.65380, 37.45220],\n",
+    "        [126.65250, 37.45180],\n",
+    "        [126.65168, 37.45100],  # 정문으로 닫기\n",
+    "    ]\n",
+    "    \n",
+    "    return {\n",
+    "        'elements': [{\n",
+    "            'type': 'way',\n",
+    "            'geometry': [{'lon': c[0], 'lat': c[1]} for c in coords],\n",
+    "            'tags': {'name': '인하대학교', 'amenity': 'university'}\n",
+    "        }]\n",
+    "    }\n",
+    "\n",
+    "# OpenStreetMap Overpass API를 사용하여 인하대학교 경계 추출\n",
+    "def get_inha_boundary():\n",
+    "    overpass_url = \"http://overpass-api.de/api/interpreter\"\n",
+    "    \n",
+    "    # 인하대학교를 검색하는 Overpass 쿼리 (더 정확한 필터링)\n",
+    "    overpass_query = \"\"\"\n",
+    "    [out:json][timeout:25];\n",
+    "    (\n",
+    "      way[\"name\"=\"인하대학교\"][\"amenity\"=\"university\"];\n",
+    "      relation[\"name\"=\"인하대학교\"][\"amenity\"=\"university\"];\n",
+    "    );\n",
+    "    out geom;\n",
+    "    \"\"\"\n",
+    "    \n",
+    "    try:\n",
+    "        response = requests.get(overpass_url, params={'data': overpass_query}, timeout=30)\n",
+    "        print(f\"응답 상태 코드: {response.status_code}\")\n",
+    "        \n",
+    "        if response.status_code != 200:\n",
+    "            print(f\"❌ API 오류 발생\")\n",
+    "            print(\"🔄 대체 방법을 시도합니다...\")\n",
+    "            return get_inha_boundary_alternative()\n",
+    "        \n",
+    "        data = response.json()\n",
+    "        if not data.get('elements'):\n",
+    "            print(\"⚠️ 검색 결과가 없습니다. 대체 방법을 시도합니다...\")\n",
+    "            return get_inha_boundary_alternative()\n",
+    "        return data\n",
+    "    except Exception as e:\n",
+    "        print(f\"❌ 요청 실패: {e}\")\n",
+    "        print(\"🔄 대체 방법을 시도합니다...\")\n",
+    "        return get_inha_boundary_alternative()\n",
+    "\n",
+    "# GeoJSON 형식으로 변환\n",
+    "def convert_to_geojson(osm_data):\n",
+    "    features = []\n",
+    "    \n",
+    "    print(f\"🔍 받은 요소 개수: {len(osm_data.get('elements', []))}\")\n",
+    "    \n",
+    "    for idx, element in enumerate(osm_data.get('elements', [])):\n",
+    "        tags = element.get('tags', {})\n",
+    "        print(f\"  [{idx}] type: {element['type']}, tags: {tags}\")\n",
+    "        \n",
+    "        if element['type'] == 'way' and 'geometry' in element:\n",
+    "            coords = [(node['lon'], node['lat']) for node in element['geometry']]\n",
+    "            print(f\"      좌표 개수: {len(coords)}\")\n",
+    "            \n",
+    "            if len(coords) > 5:  # 최소 5개 이상\n",
+    "                try:\n",
+    "                    polygon = Polygon(coords)\n",
+    "                    area = polygon.area\n",
+    "                    print(f\"      면적: {area:.8f}\")\n",
+    "                    \n",
+    "                    if polygon.is_valid:\n",
+    "                        features.append({\n",
+    "                            'type': 'Feature',\n",
+    "                            'geometry': polygon.__geo_interface__,\n",
+    "                            'properties': tags\n",
+    "                        })\n",
+    "                        print(f\"      ✓ 추가됨\")\n",
+    "                except Exception as e:\n",
+    "                    print(f\"      ✗ Polygon 생성 실패: {e}\")\n",
+    "        \n",
+    "        elif element['type'] == 'relation' and 'members' in element:\n",
+    "            all_coords = []\n",
+    "            for member in element['members']:\n",
+    "                if 'geometry' in member:\n",
+    "                    coords = [(node['lon'], node['lat']) for node in member['geometry']]\n",
+    "                    all_coords.extend(coords)\n",
+    "            \n",
+    "            print(f\"      relation 좌표 개수: {len(all_coords)}\")\n",
+    "            \n",
+    "            if len(all_coords) > 5:\n",
+    "                try:\n",
+    "                    polygon = Polygon(all_coords)\n",
+    "                    if polygon.is_valid:\n",
+    "                        features.append({\n",
+    "                            'type': 'Feature',\n",
+    "                            'geometry': polygon.__geo_interface__,\n",
+    "                            'properties': tags\n",
+    "                        })\n",
+    "                        print(f\"      ✓ relation 추가됨\")\n",
+    "                except Exception as e:\n",
+    "                    print(f\"      ✗ Relation polygon 실패: {e}\")\n",
+    "    \n",
+    "    print(f\"📊 총 {len(features)}개 feature 생성\")\n",
+    "    \n",
+    "    # 면적이 가장 큰 polygon 선택\n",
+    "    if features and len(features) > 1:\n",
+    "        features.sort(key=lambda f: Polygon(f['geometry']['coordinates'][0]).area, reverse=True)\n",
+    "        print(f\"   가장 큰 polygon 선택 (면적: {Polygon(features[0]['geometry']['coordinates'][0]).area:.8f})\")\n",
+    "        features = [features[0]]\n",
+    "    \n",
+    "    geojson = {\n",
+    "        'type': 'FeatureCollection',\n",
+    "        'features': features\n",
+    "    }\n",
+    "    \n",
+    "    return geojson\n",
+    "\n",
+    "# 경계 시각화 및 저장\n",
+    "def visualize_and_save(geojson_data, output_file='inha_boundary.geojson'):\n",
+    "    # GeoJSON 파일로 저장\n",
+    "    with open(output_file, 'w', encoding='utf-8') as f:\n",
+    "        json.dump(geojson_data, f, ensure_ascii=False, indent=2)\n",
+    "    print(f\"✓ GeoJSON 파일 저장 완료: {output_file}\")\n",
+    "    \n",
+    "    # GeoDataFrame으로 변환하여 시각화\n",
+    "    if geojson_data['features']:\n",
+    "        gdf = gpd.GeoDataFrame.from_features(geojson_data['features'])\n",
+    "        gdf.set_crs(epsg=4326, inplace=True)\n",
+    "        \n",
+    "        # Web Mercator 투영법으로 변환\n",
+    "        gdf_web = gdf.to_crs(epsg=3857)\n",
+    "        \n",
+    "        # 한글 폰트 설정\n",
+    "        try:\n",
+    "            font_path = font_manager.findfont(font_manager.FontProperties(family='Malgun Gothic'))\n",
+    "            plt.rcParams['font.family'] = 'Malgun Gothic'\n",
+    "        except:\n",
+    "            plt.rcParams['font.family'] = 'sans-serif'\n",
+    "        \n",
+    "        # 방법 1: 위성 이미지 (Esri WorldImagery - 최신)\n",
+    "        fig1, ax1 = plt.subplots(figsize=(15, 12))\n",
+    "        gdf_web.plot(ax=ax1, facecolor='none', edgecolor='red', linewidth=3, alpha=0.9)\n",
+    "        \n",
+    "        try:\n",
+    "            # 최신 위성 이미지 소스 시도\n",
+    "            ctx.add_basemap(ax1, source=ctx.providers.Esri.WorldImagery, alpha=0.8)\n",
+    "            print(\"✓ 위성 이미지 (Esri) 추가 완료\")\n",
+    "        except:\n",
+    "            try:\n",
+    "                # Google Satellite 대체\n",
+    "                ctx.add_basemap(ax1, source='https://mt1.google.com/vt/lyrs=s&x={x}&y={y}&z={z}', alpha=0.8)\n",
+    "                print(\"✓ 위성 이미지 (Google) 추가 완료\")\n",
+    "            except Exception as e:\n",
+    "                print(f\"⚠️ 위성 이미지 추가 실패: {e}\")\n",
+    "        \n",
+    "        ax1.set_title('인하대학교 경계 (위성)', fontsize=18, pad=20, weight='bold')\n",
+    "        ax1.axis('off')\n",
+    "        plt.tight_layout()\n",
+    "        plt.savefig('inha_boundary_satellite.png', dpi=300, bbox_inches='tight', facecolor='white')\n",
+    "        print(\"✓ 위성 이미지 저장 완료: inha_boundary_satellite.png\")\n",
+    "        plt.close(fig1)\n",
+    "        \n",
+    "        # 방법 2: 지도 스타일 (OpenStreetMap)\n",
+    "                # 방법 2: 지도 스타일 (OpenStreetMap) + 경계 밖 배경 클리핑\n",
+    "        fig2, ax2 = plt.subplots(figsize=(15, 12))\n",
+    "\n",
+    "        # Web Mercator 기준 폴리곤 준비 (합쳐진 하나의 geometry)\n",
+    "        poly = gdf_web.geometry.unary_union\n",
+    "\n",
+    "        # MultiPolygon이면 가장 큰 것 하나만 사용 (선택)\n",
+    "        if poly.geom_type == \"MultiPolygon\":\n",
+    "            poly = max(poly.geoms, key=lambda g: g.area)\n",
+    "\n",
+    "        # 폴리곤을 matplotlib PathPatch로 변환 (clip용)\n",
+    "        x, y = poly.exterior.coords.xy\n",
+    "        coords = np.vstack([x, y]).T\n",
+    "\n",
+    "        Path = mpath.Path\n",
+    "        path_data = [(Path.MOVETO, coords[0])]\n",
+    "        for c in coords[1:]:\n",
+    "            path_data.append((Path.LINETO, c))\n",
+    "        path_data.append((Path.CLOSEPOLY, coords[0]))\n",
+    "\n",
+    "        codes, verts = zip(*path_data)\n",
+    "        clip_path = mpath.Path(verts, codes)\n",
+    "        clip_patch = mpatches.PathPatch(clip_path, facecolor=\"none\", edgecolor=\"none\")\n",
+    "\n",
+    "        ax2.add_patch(clip_patch)\n",
+    "\n",
+    "        # 우선 폴리곤을 색칠\n",
+    "        gdf_web.plot(ax=ax2, facecolor='lightblue', edgecolor='darkblue', linewidth=2.5, alpha=0.6)\n",
+    "\n",
+    "        try:\n",
+    "            ctx.add_basemap(ax2, source=ctx.providers.OpenStreetMap.Mapnik, alpha=0.7)\n",
+    "            # 방금 추가된 배경 이미지에 클립 적용\n",
+    "            if ax2.images:\n",
+    "                ax2.images[-1].set_clip_path(clip_patch)\n",
+    "            print(\"✓ 지도 스타일 + 클리핑 추가 완료\")\n",
+    "        except Exception as e:\n",
+    "            print(f\"⚠️ 지도 추가 실패: {e}\")\n",
+    "        \n",
+    "        ax2.set_title('인하대학교 경계 (지도, boundary 밖 배경 제거)', fontsize=18, pad=20, weight='bold')\n",
+    "        ax2.axis('off')\n",
+    "        plt.tight_layout()\n",
+    "        plt.savefig('inha_boundary_map.png', dpi=300, bbox_inches='tight', facecolor='white')\n",
+    "        print(\"✓ 지도 이미지 저장 완료: inha_boundary_map.png\")\n",
+    "        plt.close(fig2)\n",
+    "\n",
+    "        \n",
+    "        # 방법 3: 경계만 (폴리곤만 강조)\n",
+    "        fig3, ax3 = plt.subplots(figsize=(12, 10))\n",
+    "        gdf.plot(ax=ax3, facecolor='#4CAF50', edgecolor='#1B5E20', linewidth=2.5, alpha=0.4)\n",
+    "        ax3.set_title('인하대학교 캠퍼스 영역', fontsize=18, pad=20, weight='bold')\n",
+    "        ax3.set_xlabel('경도 (Longitude)', fontsize=12)\n",
+    "        ax3.set_ylabel('위도 (Latitude)', fontsize=12)\n",
+    "        ax3.grid(True, alpha=0.3, linestyle='--', linewidth=0.5)\n",
+    "        ax3.set_facecolor('#f5f5f5')\n",
+    "        plt.tight_layout()\n",
+    "        plt.savefig('inha_boundary_polygon.png', dpi=300, bbox_inches='tight', facecolor='white')\n",
+    "        print(\"✓ 폴리곤 이미지 저장 완료: inha_boundary_polygon.png\")\n",
+    "        plt.close(fig3)\n",
+    "        \n",
+    "        # Shapefile로도 저장 (GIS 소프트웨어에서 사용 가능)\n",
+    "        try:\n",
+    "            gdf.to_file('inha_boundary.shp')\n",
+    "            print(\"✓ Shapefile 저장 완료: inha_boundary.shp\")\n",
+    "        except Exception as e:\n",
+    "            print(f\"⚠️ Shapefile 저장 실패: {e}\")\n",
+    "        \n",
+    "        # 경계 정보 출력\n",
+    "        print(f\"\\n📍 경계 정보:\")\n",
+    "        bounds = gdf.total_bounds\n",
+    "        print(f\"   - 서쪽: {bounds[0]:.6f}, 남쪽: {bounds[1]:.6f}\")\n",
+    "        print(f\"   - 동쪽: {bounds[2]:.6f}, 북쪽: {bounds[3]:.6f}\")\n",
+    "        \n",
+    "        # CRS 변환 후 중심점 계산 (경고 방지)\n",
+    "        gdf_projected = gdf.to_crs(epsg=5179)  # 한국 평면직각좌표계\n",
+    "        centroid_proj = gdf_projected.centroid.iloc[0]\n",
+    "        centroid_wgs84 = gpd.GeoSeries([centroid_proj], crs=5179).to_crs(4326).iloc[0]\n",
+    "        print(f\"   - 중심점: ({centroid_wgs84.x:.6f}, {centroid_wgs84.y:.6f})\")\n",
+    "        \n",
+    "        return gdf\n",
+    "    else:\n",
+    "        print(\"❌ 경계 데이터를 찾을 수 없습니다.\")\n",
+    "        return None\n",
+    "\n",
+    "# 메인 실행\n",
+    "if __name__ == \"__main__\":\n",
+    "    print(\"🔍 인하대학교 경계 정보 추출 중...\")\n",
+    "    \n",
+    "    # 1. OSM 데이터 가져오기\n",
+    "    osm_data = get_inha_boundary()\n",
+    "    print(f\"✓ {len(osm_data.get('elements', []))}개의 요소 발견\")\n",
+    "    \n",
+    "    # 2. GeoJSON으로 변환\n",
+    "    geojson = convert_to_geojson(osm_data)\n",
+    "    \n",
+    "    # 3. 저장 및 시각화\n",
+    "    gdf = visualize_and_save(geojson)\n",
+    "    \n",
+    "    # 4. 좌표 데이터 출력 (필요시)\n",
+    "    if geojson['features']:\n",
+    "        print(\"\\n📌 첫 번째 경계의 좌표:\")\n",
+    "        coords = geojson['features'][0]['geometry']['coordinates']\n",
+    "        print(f\"   총 {len(coords[0]) if coords else 0}개의 좌표점\")\n",
+    "        if coords and len(coords[0]) > 0:\n",
+    "            print(f\"   시작점: {coords[0][0]}\")\n",
+    "            print(f\"   끝점: {coords[0][-1]}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "79139ad6",
+   "metadata": {},
+   "source": [
+    "# 파일로 ��러오기"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 21,
+   "id": "6f2d7463",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Users\\USER\\AppData\\Local\\Temp\\ipykernel_4744\\2019802498.py:19: DeprecationWarning: The 'unary_union' attribute is deprecated, use the 'union_all()' method instead.\n",
+      "  poly = gdf_web.geometry.unary_union\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "✓ 저장 완료: inha_boundary_map_from_shp.png\n"
+     ]
+    }
+   ],
+   "source": [
+    "import geopandas as gpd\n",
+    "import matplotlib.pyplot as plt\n",
+    "import contextily as ctx\n",
+    "import numpy as np\n",
+    "import matplotlib.path as mpath\n",
+    "import matplotlib.patches as mpatches\n",
+    "\n",
+    "def plot_clipped_map_from_shp(\n",
+    "    shp_path=\"inha_boundary.shp\",\n",
+    "    out_png=\"inha_boundary_map_from_shp.png\"\n",
+    "):\n",
+    "    # 1) Shapefile 읽기 (shx/dbf/prj는 같이 자동으로 읽힘)\n",
+    "    gdf = gpd.read_file(shp_path)\n",
+    "\n",
+    "    # 2) Web Mercator로 변환 (contextily용)\n",
+    "    gdf_web = gdf.to_crs(epsg=3857)\n",
+    "\n",
+    "    # 3) 폴리곤 하나로 합치기 (MultiPolygon이면 가장 큰 것 선택)\n",
+    "    poly = gdf_web.geometry.unary_union\n",
+    "    if poly.geom_type == \"MultiPolygon\":\n",
+    "        poly = max(poly.geoms, key=lambda g: g.area)\n",
+    "\n",
+    "    # 4) clip에 쓸 PathPatch 만들기\n",
+    "    x, y = poly.exterior.coords.xy\n",
+    "    coords = np.vstack([x, y]).T\n",
+    "\n",
+    "    Path = mpath.Path\n",
+    "    path_data = [(Path.MOVETO, coords[0])]\n",
+    "    for c in coords[1:]:\n",
+    "        path_data.append((Path.LINETO, c))\n",
+    "    path_data.append((Path.CLOSEPOLY, coords[0]))\n",
+    "\n",
+    "    codes, verts = zip(*path_data)\n",
+    "    clip_path = mpath.Path(verts, codes)\n",
+    "    clip_patch = mpatches.PathPatch(clip_path, facecolor=\"none\", edgecolor=\"none\")\n",
+    "\n",
+    "    # 5) 그림 생성\n",
+    "    fig, ax = plt.subplots(figsize=(15, 12))\n",
+    "\n",
+    "    # 폴리곤 먼저 색칠 (경계 확인용)\n",
+    "    gdf_web.plot(ax=ax, facecolor=\"lightblue\", edgecolor=\"darkblue\", alpha=0.6, linewidth=2)\n",
+    "\n",
+    "    # clip 패치 추가\n",
+    "    ax.add_patch(clip_patch)\n",
+    "\n",
+    "    # 6) basemap 추가 + 클리핑\n",
+    "    ctx.add_basemap(ax, source=ctx.providers.OpenStreetMap.Mapnik, alpha=0.7)\n",
+    "    if ax.images:\n",
+    "        ax.images[-1].set_clip_path(clip_patch)\n",
+    "\n",
+    "    # 7) 캠퍼스 영역만 보이게 축 범위도 폴리곤에 맞게 잘라주기\n",
+    "    minx, miny, maxx, maxy = poly.bounds\n",
+    "    ax.set_xlim(minx, maxx)\n",
+    "    ax.set_ylim(miny, maxy)\n",
+    "    ax.axis(\"off\")\n",
+    "    plt.tight_layout()\n",
+    "    plt.savefig(out_png, dpi=300, bbox_inches=\"tight\", facecolor=\"white\")\n",
+    "    plt.close(fig)\n",
+    "\n",
+    "    print(f\"✓ 저장 완료: {out_png}\")\n",
+    "\n",
+    "\n",
+    "plot_clipped_map_from_shp()\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "9ee6be64",
+   "metadata": {},
+   "source": [
+    "# 표본 추출 비교"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 29,
+   "id": "c68604ed",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Users\\USER\\AppData\\Local\\Temp\\ipykernel_4744\\3417582304.py:15: DeprecationWarning: The 'unary_union' attribute is deprecated, use the 'union_all()' method instead.\n",
+      "  poly = gdf.geometry.unary_union        # 하나의 폴리곤으로 합치기\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "격자 층별 표본 수 할당 (면적비례): [ 4 12  7 12 16  9]\n",
+      "동일면적 층별 표본 수 할당 (동일 표본수): [10 10 10 10 10 10]\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Users\\USER\\AppData\\Local\\Temp\\ipykernel_4744\\3417582304.py:205: DeprecationWarning: The 'unary_union' attribute is deprecated, use the 'union_all()' method instead.\n",
+      "  poly_web = gdf_web.geometry.unary_union\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "✓ 원본 경계 지도 저장: inha_boundary_base.png\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Users\\USER\\AppData\\Local\\Temp\\ipykernel_4744\\3417582304.py:205: DeprecationWarning: The 'unary_union' attribute is deprecated, use the 'union_all()' method instead.\n",
+      "  poly_web = gdf_web.geometry.unary_union\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "✓ SRS 샘플 지도 저장: inha_sampling_srs.png\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Users\\USER\\AppData\\Local\\Temp\\ipykernel_4744\\3417582304.py:205: DeprecationWarning: The 'unary_union' attribute is deprecated, use the 'union_all()' method instead.\n",
+      "  poly_web = gdf_web.geometry.unary_union\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "✓ 층화 샘플 지도 저장: inha_sampling_strata.png\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Users\\USER\\AppData\\Local\\Temp\\ipykernel_4744\\3417582304.py:205: DeprecationWarning: The 'unary_union' attribute is deprecated, use the 'union_all()' method instead.\n",
+      "  poly_web = gdf_web.geometry.unary_union\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "✓ 동일면적 층화 샘플 지도 저장: inha_sampling_strata_equalarea.png\n",
+      "\n",
+      "[SRS 샘플 좌표 (EPSG:5179)]\n",
+      "  design                        geometry\n",
+      "0    SRS  POINT (925278.467 1939423.964)\n",
+      "1    SRS  POINT (925060.691 1939262.149)\n",
+      "2    SRS    POINT (925344.356 1939270.6)\n",
+      "3    SRS  POINT (925259.687 1939296.418)\n",
+      "4    SRS  POINT (925386.793 1939152.443)\n",
+      "\n",
+      "[격자 층화 샘플 좌표 (EPSG:5179)]\n",
+      "   stratum_id                        geometry\n",
+      "0           0   POINT (924882.652 1939247.19)\n",
+      "1           0  POINT (924924.472 1939181.424)\n",
+      "2           0   POINT (924947.85 1939212.952)\n",
+      "3           0  POINT (924930.148 1939196.918)\n",
+      "4           1  POINT (925147.586 1939174.715)\n",
+      "\n",
+      "[동일면적 층화 샘플 좌표 (EPSG:5179)]\n",
+      "   stratum_id                        geometry\n",
+      "0           0   POINT (924846.854 1939231.61)\n",
+      "1           0    POINT (924830.9 1939503.537)\n",
+      "2           0  POINT (924832.812 1939290.495)\n",
+      "3           0  POINT (924882.953 1939275.119)\n",
+      "4           0  POINT (924776.295 1939433.993)\n"
+     ]
+    }
+   ],
+   "source": [
+    "import numpy as np\n",
+    "import geopandas as gpd\n",
+    "from shapely.geometry import Point, box\n",
+    "import matplotlib.pyplot as plt\n",
+    "import contextily as ctx\n",
+    "import matplotlib.path as mpath\n",
+    "import matplotlib.patches as mpatches\n",
+    "\n",
+    "# ===========================\n",
+    "# 1. 공통: 경계 불러오기\n",
+    "# ===========================\n",
+    "def load_inha_boundary(shp_path=\"inha_boundary.shp\"):\n",
+    "    gdf = gpd.read_file(shp_path)          # prj 기준으로 CRS 인식\n",
+    "    gdf = gdf.to_crs(epsg=5179)            # 한국 평면직각좌표계\n",
+    "    poly = gdf.geometry.unary_union        # 하나의 폴리곤으로 합치기\n",
+    "    return gdf, poly\n",
+    "\n",
+    "\n",
+    "# ===========================\n",
+    "# 2. 폴리곤 내부 랜덤 샘플링 (SRS)\n",
+    "# ===========================\n",
+    "def random_points_in_polygon(poly, n):\n",
+    "    minx, miny, maxx, maxy = poly.bounds\n",
+    "    points = []\n",
+    "    # rejection sampling: bounding box에서 뽑아서, 폴리곤 안만 사용\n",
+    "    while len(points) < n:\n",
+    "        xs = np.random.uniform(minx, maxx, n * 2)\n",
+    "        ys = np.random.uniform(miny, maxy, n * 2)\n",
+    "        for x, y in zip(xs, ys):\n",
+    "            p = Point(x, y)\n",
+    "            if poly.contains(p):  # 경계까지 포함하려면 contains 대신 covers 써도 됨\n",
+    "                points.append(p)\n",
+    "                if len(points) == n:\n",
+    "                    break\n",
+    "    return points\n",
+    "\n",
+    "\n",
+    "# ===========================\n",
+    "# 3-1. 폴리곤을 격자로 나눈 후, 층(구역) 만들기 (기존 방식)\n",
+    "# ===========================\n",
+    "def make_strata(poly, n_rows=2, n_cols=3):\n",
+    "    \"\"\"\n",
+    "    n_rows x n_cols 격자 생성 후,\n",
+    "    각 cell과 캠퍼스 폴리곤의 교차(intersection)를 strata로 사용\n",
+    "    \"\"\"\n",
+    "    minx, miny, maxx, maxy = poly.bounds\n",
+    "    dx = (maxx - minx) / n_cols\n",
+    "    dy = (maxy - miny) / n_rows\n",
+    "\n",
+    "    strata_geoms = []\n",
+    "    for i in range(n_rows):\n",
+    "        for j in range(n_cols):\n",
+    "            cell = box(\n",
+    "                minx + j * dx,\n",
+    "                miny + i * dy,\n",
+    "                minx + (j + 1) * dx,\n",
+    "                miny + (i + 1) * dy,\n",
+    "            )\n",
+    "            inter = cell.intersection(poly)\n",
+    "            if not inter.is_empty and inter.area > 0:\n",
+    "                strata_geoms.append(inter)\n",
+    "\n",
+    "    strata_gdf = gpd.GeoDataFrame(\n",
+    "        {\"stratum_id\": range(len(strata_geoms))},\n",
+    "        geometry=strata_geoms,\n",
+    "        crs=\"EPSG:5179\"\n",
+    "    )\n",
+    "    return strata_gdf\n",
+    "\n",
+    "\n",
+    "# ===========================\n",
+    "# 3-2. 동일면적으로 나누는 층(구역) 만들기\n",
+    "#   - x축 방향으로 잘라서 각 밴드의 폴리곤 면적이 거의 같게\n",
+    "# ===========================\n",
+    "# ===========================\n",
+    "# 3-2. 동일면적으로 나누는 층(구역) 만들기 (빠른 근사 버전)\n",
+    "#   - 폴리곤 안에 무작위 점 많이 뽑고\n",
+    "#   - x좌표 기준으로 6등분(quantile) → 거의 동일면적 스트립\n",
+    "# ===========================\n",
+    "def make_equal_area_strata(poly, n_strata=6, n_mc=10000):\n",
+    "    \"\"\"\n",
+    "    poly를 x축 방향 스트립  n_strata개로 나누되,\n",
+    "    폴리곤 내부에 균일 샘플링한 점들의 x좌표 분포를 이용해\n",
+    "    '근사 동일면적'이 되도록 자른다.\n",
+    "    - n_mc: 몬테카를로 포인트 개수 (많을수록 면적 근사 정확, 느려짐)\n",
+    "    \"\"\"\n",
+    "\n",
+    "    # 1) 폴리곤 내부에 몬테카를로 포인트 생성\n",
+    "    mc_points = random_points_in_polygon(poly, n_mc)\n",
+    "    xs = np.array([p.x for p in mc_points])\n",
+    "\n",
+    "    # 2) x좌표 정렬 후, n_strata등분 quantile 위치 찾기\n",
+    "    xs_sorted = np.sort(xs)\n",
+    "    cuts = []\n",
+    "    for i in range(1, n_strata):\n",
+    "        q = i / n_strata\n",
+    "        idx = int(q * len(xs_sorted))\n",
+    "        cuts.append(xs_sorted[idx])\n",
+    "\n",
+    "    # 3) 잘라낼 x 경계 만들기 (좌/우 끝 포함)\n",
+    "    minx, miny, maxx, maxy = poly.bounds\n",
+    "    x_edges = [minx] + cuts + [maxx]\n",
+    "\n",
+    "    # 4) 각 스트립과 폴리곤의 교차로 strata 생성 (intersection은 스트립 개수만큼 딱 6번)\n",
+    "    strata_geoms = []\n",
+    "    for j in range(n_strata):\n",
+    "        cell = box(x_edges[j], miny, x_edges[j + 1], maxy)\n",
+    "        inter = poly.intersection(cell)\n",
+    "        if not inter.is_empty and inter.area > 0:\n",
+    "            strata_geoms.append(inter)\n",
+    "\n",
+    "    strata_gdf = gpd.GeoDataFrame(\n",
+    "        {\"stratum_id\": range(len(strata_geoms))},\n",
+    "        geometry=strata_geoms,\n",
+    "        crs=\"EPSG:5179\"\n",
+    "    )\n",
+    "    return strata_gdf\n",
+    "\n",
+    "\n",
+    "\n",
+    "\n",
+    "# ===========================\n",
+    "# 4-1. 층화 샘플링 (면적 비례 할당, 기존)\n",
+    "# ===========================\n",
+    "def stratified_sampling(poly_strata_gdf, n_total):\n",
+    "    areas = poly_strata_gdf.geometry.area.values\n",
+    "    props = areas / areas.sum()\n",
+    "\n",
+    "    # 면적 비례 초기 할당\n",
+    "    n_alloc = np.round(props * n_total).astype(int)\n",
+    "\n",
+    "    # 총합 보정 (sum == n_total 되도록)\n",
+    "    diff = n_total - n_alloc.sum()\n",
+    "    if diff > 0:\n",
+    "        # 더 필요한 만큼, 면적이 큰 층부터 1씩 추가\n",
+    "        idx_order = np.argsort(-areas)\n",
+    "        for k in range(diff):\n",
+    "            n_alloc[idx_order[k % len(n_alloc)]] += 1\n",
+    "    elif diff < 0:\n",
+    "        # 너무 많이 할당된 경우, 면적이 작은 층부터 1씩 감소 (0 밑으로는 안 가게)\n",
+    "        idx_order = np.argsort(areas)\n",
+    "        k = 0\n",
+    "        while diff < 0:\n",
+    "            i = idx_order[k % len(n_alloc)]\n",
+    "            if n_alloc[i] > 1:    # 최소 1개는 유지\n",
+    "                n_alloc[i] -= 1\n",
+    "                diff += 1\n",
+    "            k += 1\n",
+    "\n",
+    "    # 각 층에서 샘플링\n",
+    "    all_points = []\n",
+    "    stratum_ids = []\n",
+    "    for geom, sid, n_i in zip(poly_strata_gdf.geometry,\n",
+    "                              poly_strata_gdf[\"stratum_id\"],\n",
+    "                              n_alloc):\n",
+    "        pts = random_points_in_polygon(geom, n_i)\n",
+    "        all_points.extend(pts)\n",
+    "        stratum_ids.extend([sid] * len(pts))\n",
+    "\n",
+    "    sample_gdf = gpd.GeoDataFrame(\n",
+    "        {\"stratum_id\": stratum_ids},\n",
+    "        geometry=all_points,\n",
+    "        crs=poly_strata_gdf.crs,\n",
+    "    )\n",
+    "    return sample_gdf, n_alloc\n",
+    "\n",
+    "\n",
+    "# ===========================\n",
+    "# 4-2. 층화 샘플링 (동일 표본수 할당: n_total을 층 개수로 균등 분배)\n",
+    "# ===========================\n",
+    "def stratified_sampling_equal_n(poly_strata_gdf, n_total):\n",
+    "    n_strata = len(poly_strata_gdf)\n",
+    "    base = n_total // n_strata\n",
+    "    rem = n_total % n_strata\n",
+    "\n",
+    "    n_alloc = np.array([base] * n_strata)\n",
+    "    for i in range(rem):\n",
+    "        n_alloc[i] += 1\n",
+    "\n",
+    "    all_points = []\n",
+    "    stratum_ids = []\n",
+    "    for geom, sid, n_i in zip(poly_strata_gdf.geometry,\n",
+    "                              poly_strata_gdf[\"stratum_id\"],\n",
+    "                              n_alloc):\n",
+    "        pts = random_points_in_polygon(geom, n_i)\n",
+    "        all_points.extend(pts)\n",
+    "        stratum_ids.extend([sid] * len(pts))\n",
+    "\n",
+    "    sample_gdf = gpd.GeoDataFrame(\n",
+    "        {\"stratum_id\": stratum_ids},\n",
+    "        geometry=all_points,\n",
+    "        crs=poly_strata_gdf.crs,\n",
+    "    )\n",
+    "    return sample_gdf, n_alloc\n",
+    "\n",
+    "\n",
+    "# ===========================\n",
+    "# 5. 클리핑용 패치 생성\n",
+    "# ===========================\n",
+    "def make_clip_patch_and_poly_web(boundary_gdf_5179):\n",
+    "    \"\"\"\n",
+    "    경계 gdf(EPGS:5179)에서 Web Mercator(3857) 기준 폴리곤 + clip용 PathPatch 반환\n",
+    "    \"\"\"\n",
+    "    gdf_web = boundary_gdf_5179.to_crs(epsg=3857)\n",
+    "    poly_web = gdf_web.geometry.unary_union\n",
+    "\n",
+    "    # MultiPolygon이면 가장 큰 것 사용\n",
+    "    if poly_web.geom_type == \"MultiPolygon\":\n",
+    "        poly_web = max(poly_web.geoms, key=lambda g: g.area)\n",
+    "\n",
+    "    x, y = poly_web.exterior.coords.xy\n",
+    "    coords = np.vstack([x, y]).T\n",
+    "\n",
+    "    Path = mpath.Path\n",
+    "    path_data = [(Path.MOVETO, coords[0])]\n",
+    "    for c in coords[1:]:\n",
+    "        path_data.append((Path.LINETO, c))\n",
+    "    path_data.append((Path.CLOSEPOLY, coords[0]))\n",
+    "\n",
+    "    codes, verts = zip(*path_data)\n",
+    "    clip_path = mpath.Path(verts, codes)\n",
+    "    clip_patch = mpatches.PathPatch(\n",
+    "        clip_path,\n",
+    "        facecolor=\"none\",\n",
+    "        edgecolor=\"none\"\n",
+    "    )\n",
+    "\n",
+    "    return clip_patch, poly_web, gdf_web\n",
+    "\n",
+    "\n",
+    "# ===========================\n",
+    "# 6. 시각화 1: 원본 경계만 (샘플 없음)\n",
+    "# ===========================\n",
+    "def plot_base(gdf_boundary_5179,\n",
+    "              out_png=\"inha_boundary_base.png\"):\n",
+    "\n",
+    "    clip_patch, poly_web, boundary_web = make_clip_patch_and_poly_web(gdf_boundary_5179)\n",
+    "\n",
+    "    fig, ax = plt.subplots(figsize=(12, 10))\n",
+    "\n",
+    "    # clip 패치\n",
+    "    ax.add_patch(clip_patch)\n",
+    "\n",
+    "    # 경계(선)\n",
+    "    boundary_web.boundary.plot(ax=ax, linewidth=2, edgecolor=\"black\")\n",
+    "\n",
+    "    # basemap\n",
+    "    ctx.add_basemap(ax, source=ctx.providers.OpenStreetMap.Mapnik, alpha=0.9)\n",
+    "\n",
+    "    # basemap 클리핑\n",
+    "    if ax.images:\n",
+    "        ax.images[-1].set_clip_path(clip_patch)\n",
+    "\n",
+    "    # 축 범위\n",
+    "    minx, miny, maxx, maxy = poly_web.bounds\n",
+    "    ax.set_xlim(minx, maxx)\n",
+    "    ax.set_ylim(miny, maxy)\n",
+    "\n",
+    "    # 바깥 투명\n",
+    "    ax.set_facecolor(\"none\")\n",
+    "    fig.patch.set_alpha(0.0)\n",
+    "\n",
+    "    ax.axis(\"off\")\n",
+    "    plt.tight_layout()\n",
+    "    plt.savefig(out_png, dpi=300, bbox_inches=\"tight\", transparent=True)\n",
+    "    plt.close(fig)\n",
+    "    print(f\"✓ 원본 경계 지도 저장: {out_png}\")\n",
+    "\n",
+    "\n",
+    "# ===========================\n",
+    "# 7. 시각화 2: SRS 샘플\n",
+    "# ===========================\n",
+    "def plot_srs(gdf_boundary_5179, srs_points_5179,\n",
+    "             out_png=\"inha_sampling_srs.png\"):\n",
+    "\n",
+    "    clip_patch, poly_web, boundary_web = make_clip_patch_and_poly_web(gdf_boundary_5179)\n",
+    "    pts_web = srs_points_5179.to_crs(epsg=3857)\n",
+    "\n",
+    "    fig, ax = plt.subplots(figsize=(12, 10))\n",
+    "\n",
+    "    ax.add_patch(clip_patch)\n",
+    "\n",
+    "    # 경계(선)\n",
+    "    boundary_web.boundary.plot(ax=ax, linewidth=2, edgecolor=\"black\")\n",
+    "\n",
+    "    # SRS 점\n",
+    "    pts_web.plot(ax=ax, markersize=30, color=\"red\", alpha=0.9)\n",
+    "\n",
+    "    # basemap\n",
+    "    ctx.add_basemap(ax, source=ctx.providers.OpenStreetMap.Mapnik, alpha=0.9)\n",
+    "\n",
+    "    # basemap 클리핑\n",
+    "    if ax.images:\n",
+    "        ax.images[-1].set_clip_path(clip_patch)\n",
+    "\n",
+    "    # 축 범위\n",
+    "    minx, miny, maxx, maxy = poly_web.bounds\n",
+    "    ax.set_xlim(minx, maxx)\n",
+    "    ax.set_ylim(miny, maxy)\n",
+    "\n",
+    "    # 바깥 투명\n",
+    "    ax.set_facecolor(\"none\")\n",
+    "    fig.patch.set_alpha(0.0)\n",
+    "\n",
+    "    ax.axis(\"off\")\n",
+    "    plt.tight_layout()\n",
+    "    plt.savefig(out_png, dpi=300, bbox_inches=\"tight\", transparent=True)\n",
+    "    plt.close(fig)\n",
+    "    print(f\"✓ SRS 샘플 지도 저장: {out_png}\")\n",
+    "\n",
+    "\n",
+    "# ===========================\n",
+    "# 8-1. 시각화 3: 층화 샘플 (격자 + 면적��례)\n",
+    "# ===========================\n",
+    "def plot_strata(gdf_boundary_5179, strata_gdf_5179, sample_gdf_5179,\n",
+    "                out_png=\"inha_sampling_strata.png\"):\n",
+    "\n",
+    "    clip_patch, poly_web, boundary_web = make_clip_patch_and_poly_web(gdf_boundary_5179)\n",
+    "    strata_web = strata_gdf_5179.to_crs(epsg=3857)\n",
+    "    sample_web = sample_gdf_5179.to_crs(epsg=3857)\n",
+    "\n",
+    "    fig, ax = plt.subplots(figsize=(12, 10))\n",
+    "\n",
+    "    ax.add_patch(clip_patch)\n",
+    "\n",
+    "    # 층(구역) 영역 색칠\n",
+    "    strata_web.plot(\n",
+    "        ax=ax,\n",
+    "        column=\"stratum_id\",\n",
+    "        alpha=0.35,\n",
+    "        edgecolor=\"black\",\n",
+    "        linewidth=1,\n",
+    "        legend=False\n",
+    "    )\n",
+    "\n",
+    "    # 캠퍼스 전체 경계\n",
+    "    boundary_web.boundary.plot(ax=ax, linewidth=2, edgecolor=\"black\")\n",
+    "\n",
+    "    # 층화 샘플 점 (층별 색 구분)\n",
+    "    sample_web.plot(\n",
+    "        ax=ax,\n",
+    "        column=\"stratum_id\",\n",
+    "        markersize=30,\n",
+    "        cmap=\"tab10\",\n",
+    "        legend=False\n",
+    "    )\n",
+    "\n",
+    "    # basemap\n",
+    "    ctx.add_basemap(ax, source=ctx.providers.OpenStreetMap.Mapnik, alpha=0.9)\n",
+    "\n",
+    "    # basemap 클리핑\n",
+    "    if ax.images:\n",
+    "        ax.images[-1].set_clip_path(clip_patch)\n",
+    "\n",
+    "    # 축 범위\n",
+    "    minx, miny, maxx, maxy = poly_web.bounds\n",
+    "    ax.set_xlim(minx, maxx)\n",
+    "    ax.set_ylim(miny, maxy)\n",
+    "\n",
+    "    # 바깥 투명\n",
+    "    ax.set_facecolor(\"none\")\n",
+    "    fig.patch.set_alpha(0.0)\n",
+    "\n",
+    "    ax.axis(\"off\")\n",
+    "    plt.tight_layout()\n",
+    "    plt.savefig(out_png, dpi=300, bbox_inches=\"tight\", transparent=True)\n",
+    "    plt.close(fig)\n",
+    "    print(f\"✓ 층화 샘플 지도 저장: {out_png}\")\n",
+    "\n",
+    "\n",
+    "# ===========================\n",
+    "# 8-2. 시각화 4: 층화 샘플 (동일면적 strata + 동일 표본수)\n",
+    "# ===========================\n",
+    "def plot_strata_equal(gdf_boundary_5179, strata_gdf_5179, sample_gdf_5179,\n",
+    "                      out_png=\"inha_sampling_strata_equalarea.png\"):\n",
+    "\n",
+    "    clip_patch, poly_web, boundary_web = make_clip_patch_and_poly_web(gdf_boundary_5179)\n",
+    "    strata_web = strata_gdf_5179.to_crs(epsg=3857)\n",
+    "    sample_web = sample_gdf_5179.to_crs(epsg=3857)\n",
+    "\n",
+    "    fig, ax = plt.subplots(figsize=(12, 10))\n",
+    "\n",
+    "    ax.add_patch(clip_patch)\n",
+    "\n",
+    "    # 층(구역) 영역 색칠\n",
+    "    strata_web.plot(\n",
+    "        ax=ax,\n",
+    "        column=\"stratum_id\",\n",
+    "        alpha=0.35,\n",
+    "        edgecolor=\"black\",\n",
+    "        linewidth=1,\n",
+    "        legend=False\n",
+    "    )\n",
+    "\n",
+    "    # 캠퍼스 전체 경계\n",
+    "    boundary_web.boundary.plot(ax=ax, linewidth=2, edgecolor=\"black\")\n",
+    "\n",
+    "    # 층화 샘플 점 (층별 색 구분)\n",
+    "    sample_web.plot(\n",
+    "        ax=ax,\n",
+    "        column=\"stratum_id\",\n",
+    "        markersize=30,\n",
+    "        cmap=\"tab10\",\n",
+    "        legend=False\n",
+    "    )\n",
+    "\n",
+    "    # basemap\n",
+    "    ctx.add_basemap(ax, source=ctx.providers.OpenStreetMap.Mapnik, alpha=0.9)\n",
+    "\n",
+    "    # basemap 클리핑\n",
+    "    if ax.images:\n",
+    "        ax.images[-1].set_clip_path(clip_patch)\n",
+    "\n",
+    "    # 축 범위\n",
+    "    minx, miny, maxx, maxy = poly_web.bounds\n",
+    "    ax.set_xlim(minx, maxx)\n",
+    "    ax.set_ylim(miny, maxy)\n",
+    "\n",
+    "    # 바깥 투명\n",
+    "    ax.set_facecolor(\"none\")\n",
+    "    fig.patch.set_alpha(0.0)\n",
+    "\n",
+    "\n",
+    "    ax.axis(\"off\")\n",
+    "    plt.tight_layout()\n",
+    "    plt.savefig(out_png, dpi=300, bbox_inches=\"tight\", transparent=True)\n",
+    "    plt.close(fig)\n",
+    "    print(f\"✓ 동일면적 층화 샘플 지도 저장: {out_png}\")\n",
+    "\n",
+    "\n",
+    "# ===========================\n",
+    "# 9. 메인 실행\n",
+    "# ===========================\n",
+    "if __name__ == \"__main__\":\n",
+    "    # 1) 경계 불러오기\n",
+    "    boundary_gdf, boundary_poly = load_inha_boundary(\"inha_boundary.shp\")\n",
+    "\n",
+    "    # 2) SRS 샘플링 (예: 40개)\n",
+    "    n_srs = 60\n",
+    "    srs_points = random_points_in_polygon(boundary_poly, n_srs)\n",
+    "    srs_gdf = gpd.GeoDataFrame(\n",
+    "        {\"design\": [\"SRS\"] * len(srs_points)},\n",
+    "        geometry=srs_points,\n",
+    "        crs=\"EPSG:5179\"\n",
+    "    )\n",
+    "\n",
+    "    # 3-1) 격자 층(구역) 만들고 층화 샘플링 (면적비례)\n",
+    "    strata_gdf = make_strata(boundary_poly, n_rows=2, n_cols=3)\n",
+    "    strat_sample_gdf, n_alloc = stratified_sampling(strata_gdf, n_total=60)\n",
+    "    print(\"격자 층별 표본 수 할당 (면적비례):\", n_alloc)\n",
+    "\n",
+    "    # 3-2) 동일면적 strata 만들고 층화 샘플링 (동일 표본수)\n",
+    "    equal_strata_gdf = make_equal_area_strata(boundary_poly, n_strata=6)\n",
+    "    equal_strat_sample_gdf, n_alloc_equal = stratified_sampling_equal_n(equal_strata_gdf, n_total=60)\n",
+    "    print(\"동일면적 층별 표본 수 할당 (동일 표본수):\", n_alloc_equal)\n",
+    "\n",
+    "    # 4) 원본 경계 지도 (바깥 투명)\n",
+    "    plot_base(boundary_gdf, out_png=\"inha_boundary_base.png\")\n",
+    "\n",
+    "    # 5) SRS 지도\n",
+    "    plot_srs(boundary_gdf, srs_gdf,\n",
+    "             out_png=\"inha_sampling_srs.png\")\n",
+    "\n",
+    "    # 6) 층화 지도 (격자 + 면적비례)\n",
+    "    plot_strata(boundary_gdf, strata_gdf, strat_sample_gdf,\n",
+    "                out_png=\"inha_sampling_strata.png\")\n",
+    "\n",
+    "    # 7) 층화 지도 (동일면적 strata + 동일 표본수)\n",
+    "    plot_strata_equal(boundary_gdf, equal_strata_gdf, equal_strat_sample_gdf,\n",
+    "                      out_png=\"inha_sampling_strata_equalarea.png\")\n",
+    "\n",
+    "    # 8) 좌표 확인용 출력\n",
+    "    print(\"\\n[SRS 샘플 좌표 (EPSG:5179)]\")\n",
+    "    print(srs_gdf.head())\n",
+    "\n",
+    "    print(\"\\n[격자 층화 샘플 좌표 (EPSG:5179)]\")\n",
+    "    print(strat_sample_gdf.head())\n",
+    "\n",
+    "    print(\"\\n[동일면적 층화 샘플 좌표 (EPSG:5179)]\")\n",
+    "    print(equal_strat_sample_gdf.head())\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "torch_ev",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.10.16"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

데이터 수집/김다원.csv

@@ -0,0 +1,92 @@
+lat,lon,dB
+37.4516638,126.6527864,50.0
+37.4512905,126.653364,49.0
+37.4514279,126.652793,48.0
+37.4511085,126.6535225,49.0
+37.4510701,126.6531148,48.0
+37.4513171,126.652632,47.0
+37.4510595,126.6526267,50.0
+37.4510309,126.6528753,51.0
+37.4510631,126.6536844,57.0
+37.450863,126.6535503,52.0
+37.4507669,126.6537015,51.0
+37.4512012,126.6539697,47.0
+37.4505606,126.6533733,47.0
+37.4509074,126.6521297,58.0
+37.4507437,126.6524077,49.0
+37.4506799,126.6527081,48.0
+37.4505351,126.6524077,53.0
+37.4503988,126.6532875,47.0
+37.4503967,126.6529463,46.0
+37.4512256,126.6514948,47.0
+37.4510425,126.6515538,46.0
+37.451127,126.6513348,46.0
+37.4513746,126.6517523,42.0
+37.4506117,126.6531748,44.0
+37.4509056,126.6523755,42.0
+37.4517914,126.6530514,55.0
+37.4514975,126.6537273,42.0
+37.4515699,126.6534377,41.0
+37.4517164,126.6525593,39.0
+37.4514336,126.6522414,50.0
+37.4512906,126.6521779,42.0
+37.4506252,126.6513859,53.0
+37.4504676,126.6517346,45.0
+37.4502887,126.6514664,50.0
+37.4501865,126.6517293,47.0
+37.4500971,126.651917,49.0
+37.4499438,126.6512143,50.0
+37.4499054,126.6506027,51.0
+37.4503398,126.6505276,49.0
+37.4505698,126.6506188,43.0
+37.4506251,126.6509514,45.0
+37.4500532,126.6527818,46.0
+37.4499255,126.6529267,43.0
+37.4497977,126.6530554,44.0
+37.4495805,126.6526316,48.0
+37.449474,126.6528355,52.0
+37.4492398,126.6524492,53.0
+37.4489332,126.6520201,50.0
+37.4489204,126.6522293,46.0
+37.4479436,126.6529785,49.0
+37.4486122,126.6533164,46.0
+37.4486569,126.6536919,43.0
+37.4489806,126.6539548,44.0
+37.4488827,126.6542498,47.0
+37.4495172,126.6545878,38.0
+37.4494917,126.6541855,51.0
+37.4496024,126.654502,52.0
+37.4499814,126.6539816,52.0
+37.4497003,126.6544966,46.0
+37.4495939,126.654797,54.0
+37.4493171,126.6552154,46.0
+37.4491893,126.6552101,47.0
+37.4491978,126.655076,45.0
+37.4490317,126.6551242,47.0
+37.4513577,126.6545939,51.0
+37.4505187,126.6540735,53.0
+37.4511234,126.6544705,60.0
+37.4508636,126.6558974,43.0
+37.4509659,126.6549157,44.0
+37.4508935,126.6552698,39.0
+37.4505826,126.6550338,56.0
+37.4506465,126.6546851,52.0
+37.4503569,126.6549479,59.0
+37.4504548,126.6550123,54.0
+37.4505655,126.655243,42.0
+37.4503654,126.655495,47.0
+37.4505016,126.6553824,41.0
+37.4500523,126.6554818,41.0
+37.4498416,126.6553664,46.0
+37.450259,126.6542881,44.0
+37.4502206,126.6544383,43.0
+37.4500119,126.6549318,44.0
+37.449931,126.6551249,51.0
+37.4499437,126.6552644,46.0
+37.4501013,126.6545617,50.0
+37.4501525,126.654213,48.0
+37.4499771,126.65462,48.0
+37.4497514,126.6551403,52.0
+37.4498331,126.6546207,48.0
+37.4497777,126.6547173,52.0
+37.4490231,126.6561154,0.0

데이터 수집/김다원.ipynb

@@ -0,0 +1,159 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 26,
+   "id": "be7e70ab",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Users\\USER\\AppData\\Local\\Temp\\ipykernel_20792\\1924603428.py:19: DeprecationWarning: The 'unary_union' attribute is deprecated, use the 'union_all()' method instead.\n",
+      "  poly = gdf_web.geometry.unary_union\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "포인트 개수: 90\n",
+      "저장 완료: inha_boundary_with_points.png\n"
+     ]
+    }
+   ],
+   "source": [
+    "import geopandas as gpd\n",
+    "import pandas as pd\n",
+    "import matplotlib.pyplot as plt\n",
+    "import contextily as ctx\n",
+    "import numpy as np\n",
+    "import matplotlib.path as mpath\n",
+    "import matplotlib.patches as mpatches\n",
+    "from shapely import wkt\n",
+    "\n",
+    "def plot_clipped_map_with_points(\n",
+    "    shp_path=r\"C:\\Users\\USER\\iCloudDrive\\인하대학교\\5학년 1학기\\표본론 및 실습\\TeamProject\\Data\\inha_boundary.shp\",\n",
+    "    points_csv=r\"C:\\Users\\USER\\iCloudDrive\\인하대학교\\5학년 1학기\\표본론 및 실습\\TeamProject\\데이터 수집\\김다원.csv\",\n",
+    "    out_png=\"inha_boundary_with_points.png\",\n",
+    "):\n",
+    "    # ---------- 1) 캠퍼스 경계 ----------\n",
+    "    gdf = gpd.read_file(shp_path)\n",
+    "    gdf_web = gdf.to_crs(epsg=3857)\n",
+    "\n",
+    "    poly = gdf_web.geometry.unary_union\n",
+    "    if poly.geom_type == \"MultiPolygon\":\n",
+    "        poly = max(poly.geoms, key=lambda g: g.area)\n",
+    "\n",
+    "    x, y = poly.exterior.coords.xy\n",
+    "    coords = np.vstack([x, y]).T\n",
+    "\n",
+    "    Path = mpath.Path\n",
+    "    path_data = [(Path.MOVETO, coords[0])]\n",
+    "    for c in coords[1:]:\n",
+    "        path_data.append((Path.LINETO, c))\n",
+    "    path_data.append((Path.CLOSEPOLY, coords[0]))\n",
+    "\n",
+    "    codes, verts = zip(*path_data)\n",
+    "    clip_path = mpath.Path(verts, codes)\n",
+    "    clip_patch = mpatches.PathPatch(clip_path, facecolor=\"none\", edgecolor=\"none\")\n",
+    "\n",
+    "    # ---------- 2) CSV 포인트 읽기 ----------\n",
+    "    df_pts = pd.read_csv(points_csv, encoding=\"cp949\")\n",
+    "\n",
+    "    # 컬럼 이름에 BOM 붙어 있을 가능성 처리\n",
+    "    df_pts.columns = [c.strip() for c in df_pts.columns]\n",
+    "\n",
+    "    # WKT 컬럼 찾기\n",
+    "    if \"WKT\" not in df_pts.columns:\n",
+    "        # 이름에 'WKT'가 포함된 컬럼 찾기\n",
+    "        cand = [c for c in df_pts.columns if \"WKT\" in c.upper()]\n",
+    "        if not cand:\n",
+    "            # 그래도 없으면 첫 번째 컬럼을 WKT로 간주\n",
+    "            cand = [df_pts.columns[0]]\n",
+    "        df_pts = df_pts.rename(columns={cand[0]: \"WKT\"})\n",
+    "\n",
+    "    # dB 컬럼 찾기\n",
+    "    if \"dB\" not in df_pts.columns:\n",
+    "        value_cols = [c for c in df_pts.columns if c != \"WKT\"]\n",
+    "        if not value_cols:\n",
+    "            raise ValueError(\"dB 값으로 쓸 숫자 컬럼이 없음.\")\n",
+    "        df_pts = df_pts.rename(columns={value_cols[0]: \"dB\"})\n",
+    "\n",
+    "    # WKT → geometry\n",
+    "    df_pts[\"geometry\"] = df_pts[\"WKT\"].apply(wkt.loads)\n",
+    "    gdf_pts = gpd.GeoDataFrame(df_pts, geometry=\"geometry\", crs=4326)\n",
+    "    gdf_pts_web = gdf_pts.to_crs(epsg=3857)\n",
+    "\n",
+    "    print(\"포인트 개수:\", len(gdf_pts_web))\n",
+    "\n",
+    "    # ---------- 3) 그림 ----------\n",
+    "    fig, ax = plt.subplots(figsize=(15, 12))\n",
+    "\n",
+    "    # 폴리곤: 테두리만\n",
+    "    gdf_web.plot(\n",
+    "        ax=ax,\n",
+    "        facecolor=\"none\",      # 내부 색칠 X\n",
+    "        edgecolor=\"darkblue\",\n",
+    "        linewidth=2,\n",
+    "    )\n",
+    "\n",
+    "    ax.add_patch(clip_patch)\n",
+    "\n",
+    "    # 베이스맵\n",
+    "    ctx.add_basemap(ax, source=ctx.providers.OpenStreetMap.Mapnik, alpha=0.7)\n",
+    "    if ax.images:\n",
+    "        ax.images[-1].set_clip_path(clip_patch)\n",
+    "\n",
+    "    # 포인트 (컬러바 없음)\n",
+    "    gdf_pts_web.plot(\n",
+    "        ax=ax,\n",
+    "        column=\"dB\",\n",
+    "        cmap=\"viridis\",\n",
+    "        markersize=40,\n",
+    "        edgecolor=\"black\",\n",
+    "        linewidth=0.3,\n",
+    "        alpha=0.9,\n",
+    "        legend=False,   # 컬러바 제거\n",
+    "        zorder=5,       # 항상 위에 오게\n",
+    "    )\n",
+    "\n",
+    "    minx, miny, maxx, maxy = poly.bounds\n",
+    "    ax.set_xlim(minx, maxx)\n",
+    "    ax.set_ylim(miny, maxy)\n",
+    "    ax.axis(\"off\")\n",
+    "\n",
+    "    plt.tight_layout()\n",
+    "    plt.savefig(out_png, dpi=300, bbox_inches=\"tight\", facecolor=\"white\")\n",
+    "    plt.close(fig)\n",
+    "\n",
+    "    print(f\"저장 완료: {out_png}\")\n",
+    "\n",
+    "\n",
+    "plot_clipped_map_with_points()\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "torch_ev",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.10.16"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

데이터 수집/측정결과.csv

@@ -0,0 +1,114 @@
+dB
+45.0
+52.0
+55.0
+49.0
+56.0
+49.0
+37.0
+47.0
+41.0
+42.0
+59.0
+46.0
+43.0
+47.0
+53.0
+46.0
+50.0
+54.0
+44.0
+0.0
+47.0
+55.0
+47.0
+45.0
+53.0
+59.0
+58.0
+46.0
+63.0
+52.0
+51.0
+61.0
+38.0
+48.0
+54.0
+52.0
+46.0
+58.0
+52.0
+52.0
+44.0
+48.0
+46.0
+51.0
+43.0
+51.0
+56.0
+46.0
+50.0
+48.0
+52.0
+44.0
+41.0
+46.0
+49.0
+50.0
+48.0
+47.0
+43.0
+44.0
+50.0
+49.0
+59.0
+47.0
+58.0
+46.0
+47.0
+54.0
+45.0
+41.0
+53.0
+53.0
+48.0
+47.0
+42.0
+43.0
+56.0
+44.0
+45.0
+53.0
+52.0
+51.0
+48.0
+49.0
+51.0
+52.0
+43.0
+39.0
+42.0
+58.0
+44.0
+51.0
+46.0
+50.0
+57.0
+48.0
+49.0
+60.0
+46.0
+47.0
+47.0
+49.0
+42.0
+47.0
+51.0
+42.0
+48.0
+50.0
+42.0
+41.0
+50.0
+39.0
+55.0

모수탐정단_제안서.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac79521fc0e43b7f144b53c0aada854cfc8186062c676285645280c590cb709b
+size 2663276

아이디어/외접최대원.txt

@@ -0,0 +1,316 @@
+import React, { useState, useEffect, useRef } from 'react';
+
+const PolygonCircleFinder = () => {
+  const canvasRef = useRef(null);
+  const [polygon, setPolygon] = useState([
+    { x: 200, y: 100 },
+    { x: 350, y: 150 },
+    { x: 400, y: 300 },
+    { x: 250, y: 350 },
+    { x: 100, y: 250 }
+  ]);
+  const [minCircle, setMinCircle] = useState(null);
+  const [maxInnerCircle, setMaxInnerCircle] = useState(null);
+  const [draggingPoint, setDraggingPoint] = useState(null);
+
+  // 최소 외접원 찾기 (Welzl's algorithm)
+  const findMinEnclosingCircle = (points) => {
+    const shuffled = [...points].sort(() => Math.random() - 0.5);
+    return welzl(shuffled, [], shuffled.length);
+  };
+
+  const welzl = (P, R, n) => {
+    if (n === 0 || R.length === 3) {
+      return makeCircle(R);
+    }
+    
+    const p = P[n - 1];
+    const circle = welzl(P, R, n - 1);
+    
+    if (circle && isInside(circle, p)) {
+      return circle;
+    }
+    
+    return welzl(P, [...R, p], n - 1);
+  };
+
+  const makeCircle = (points) => {
+    if (points.length === 0) return null;
+    if (points.length === 1) return { x: points[0].x, y: points[0].y, r: 0 };
+    if (points.length === 2) {
+      const center = {
+        x: (points[0].x + points[1].x) / 2,
+        y: (points[0].y + points[1].y) / 2
+      };
+      const r = distance(points[0], points[1]) / 2;
+      return { ...center, r };
+    }
+    
+    // 3점으로 원 만들기
+    return circleFrom3Points(points[0], points[1], points[2]);
+  };
+
+  const circleFrom3Points = (p1, p2, p3) => {
+    const ax = p1.x, ay = p1.y;
+    const bx = p2.x, by = p2.y;
+    const cx = p3.x, cy = p3.y;
+    
+    const d = 2 * (ax * (by - cy) + bx * (cy - ay) + cx * (ay - by));
+    if (Math.abs(d) < 1e-10) return null;
+    
+    const ux = ((ax * ax + ay * ay) * (by - cy) + (bx * bx + by * by) * (cy - ay) + (cx * cx + cy * cy) * (ay - by)) / d;
+    const uy = ((ax * ax + ay * ay) * (cx - bx) + (bx * bx + by * by) * (ax - cx) + (cx * cx + cy * cy) * (bx - ax)) / d;
+    
+    const r = distance({ x: ux, y: uy }, p1);
+    return { x: ux, y: uy, r };
+  };
+
+  const distance = (p1, p2) => {
+    return Math.sqrt((p1.x - p2.x) ** 2 + (p1.y - p2.y) ** 2);
+  };
+
+  const isInside = (circle, point) => {
+    if (!circle) return false;
+    return distance(circle, point) <= circle.r + 1e-10;
+  };
+
+  // 점이 폴리곤 내부에 있는지 확인
+  const isPointInPolygon = (point, polygon) => {
+    let inside = false;
+    for (let i = 0, j = polygon.length - 1; i < polygon.length; j = i++) {
+      const xi = polygon[i].x, yi = polygon[i].y;
+      const xj = polygon[j].x, yj = polygon[j].y;
+      
+      const intersect = ((yi > point.y) !== (yj > point.y))
+        && (point.x < (xj - xi) * (point.y - yi) / (yj - yi) + xi);
+      if (intersect) inside = !inside;
+    }
+    return inside;
+  };
+
+  // 점에서 폴리곤까지의 최단 거리
+  const distanceToPolygon = (point, polygon) => {
+    let minDist = Infinity;
+    
+    for (let i = 0; i < polygon.length; i++) {
+      const j = (i + 1) % polygon.length;
+      const dist = distanceToSegment(point, polygon[i], polygon[j]);
+      minDist = Math.min(minDist, dist);
+    }
+    
+    return minDist;
+  };
+
+  const distanceToSegment = (p, v, w) => {
+    const l2 = distance(v, w) ** 2;
+    if (l2 === 0) return distance(p, v);
+    
+    let t = ((p.x - v.x) * (w.x - v.x) + (p.y - v.y) * (w.y - v.y)) / l2;
+    t = Math.max(0, Math.min(1, t));
+    
+    const projection = {
+      x: v.x + t * (w.x - v.x),
+      y: v.y + t * (w.y - v.y)
+    };
+    
+    return distance(p, projection);
+  };
+
+  // 외접원 안에서 폴리곤 외부의 최대 원 찾기
+  const findMaxInnerCircle = (enclosingCircle, polygon) => {
+    if (!enclosingCircle) return null;
+    
+    let maxCircle = null;
+    let maxRadius = 0;
+    
+    // 그리드 샘플링으로 최적 위치 찾기
+    const samples = 50;
+    const step = (enclosingCircle.r * 2) / samples;
+    
+    for (let i = 0; i <= samples; i++) {
+      for (let j = 0; j <= samples; j++) {
+        const x = enclosingCircle.x - enclosingCircle.r + i * step;
+        const y = enclosingCircle.y - enclosingCircle.r + j * step;
+        const point = { x, y };
+        
+        // 외접원 안에 있는지 확인
+        if (distance(point, enclosingCircle) > enclosingCircle.r) continue;
+        
+        // 폴리곤 외부에 있는지 확인
+        if (isPointInPolygon(point, polygon)) continue;
+        
+        // 폴리곤까지의 거리와 외접원 경계까지의 거리 중 작은 값
+        const distToPoly = distanceToPolygon(point, polygon);
+        const distToEnclosing = enclosingCircle.r - distance(point, enclosingCircle);
+        const radius = Math.min(distToPoly, distToEnclosing);
+        
+        if (radius > maxRadius) {
+          maxRadius = radius;
+          maxCircle = { x, y, r: radius };
+        }
+      }
+    }
+    
+    return maxCircle;
+  };
+
+  useEffect(() => {
+    const mec = findMinEnclosingCircle(polygon);
+    setMinCircle(mec);
+    
+    if (mec) {
+      const mic = findMaxInnerCircle(mec, polygon);
+      setMaxInnerCircle(mic);
+    }
+  }, [polygon]);
+
+  useEffect(() => {
+    const canvas = canvasRef.current;
+    if (!canvas) return;
+    
+    const ctx = canvas.getContext('2d');
+    ctx.clearRect(0, 0, canvas.width, canvas.height);
+    
+    // 최소 외접원 그리기
+    if (minCircle) {
+      ctx.strokeStyle = '#3b82f6';
+      ctx.lineWidth = 2;
+      ctx.beginPath();
+      ctx.arc(minCircle.x, minCircle.y, minCircle.r, 0, 2 * Math.PI);
+      ctx.stroke();
+      
+      ctx.fillStyle = '#3b82f6';
+      ctx.font = '14px sans-serif';
+      ctx.fillText(`외접원 r=${minCircle.r.toFixed(1)}`, minCircle.x - 60, minCircle.y - minCircle.r - 10);
+    }
+    
+    // 폴리곤 그리기
+    ctx.fillStyle = 'rgba(139, 92, 246, 0.3)';
+    ctx.strokeStyle = '#8b5cf6';
+    ctx.lineWidth = 2;
+    ctx.beginPath();
+    ctx.moveTo(polygon[0].x, polygon[0].y);
+    for (let i = 1; i < polygon.length; i++) {
+      ctx.lineTo(polygon[i].x, polygon[i].y);
+    }
+    ctx.closePath();
+    ctx.fill();
+    ctx.stroke();
+    
+    // 폴리곤 꼭짓점 그리기
+    polygon.forEach((point, i) => {
+      ctx.fillStyle = '#8b5cf6';
+      ctx.beginPath();
+      ctx.arc(point.x, point.y, 6, 0, 2 * Math.PI);
+      ctx.fill();
+    });
+    
+    // 최대 내접원 그리기
+    if (maxInnerCircle) {
+      ctx.strokeStyle = '#10b981';
+      ctx.fillStyle = 'rgba(16, 185, 129, 0.2)';
+      ctx.lineWidth = 2;
+      ctx.beginPath();
+      ctx.arc(maxInnerCircle.x, maxInnerCircle.y, maxInnerCircle.r, 0, 2 * Math.PI);
+      ctx.fill();
+      ctx.stroke();
+      
+      ctx.fillStyle = '#10b981';
+      ctx.font = '14px sans-serif';
+      ctx.fillText(`최대원 r=${maxInnerCircle.r.toFixed(1)}`, maxInnerCircle.x - 50, maxInnerCircle.y + maxInnerCircle.r + 20);
+    }
+  }, [polygon, minCircle, maxInnerCircle]);
+
+  const handleMouseDown = (e) => {
+    const canvas = canvasRef.current;
+    const rect = canvas.getBoundingClientRect();
+    const x = e.clientX - rect.left;
+    const y = e.clientY - rect.top;
+    
+    for (let i = 0; i < polygon.length; i++) {
+      if (distance({ x, y }, polygon[i]) < 10) {
+        setDraggingPoint(i);
+        break;
+      }
+    }
+  };
+
+  const handleMouseMove = (e) => {
+    if (draggingPoint === null) return;
+    
+    const canvas = canvasRef.current;
+    const rect = canvas.getBoundingClientRect();
+    const x = e.clientX - rect.left;
+    const y = e.clientY - rect.top;
+    
+    const newPolygon = [...polygon];
+    newPolygon[draggingPoint] = { x, y };
+    setPolygon(newPolygon);
+  };
+
+  const handleMouseUp = () => {
+    setDraggingPoint(null);
+  };
+
+  const resetPolygon = () => {
+    setPolygon([
+      { x: 200, y: 100 },
+      { x: 350, y: 150 },
+      { x: 400, y: 300 },
+      { x: 250, y: 350 },
+      { x: 100, y: 250 }
+    ]);
+  };
+
+  return (
+    <div className="w-full max-w-4xl mx-auto p-6 bg-gray-50 rounded-lg">
+      <h1 className="text-2xl font-bold mb-4 text-gray-800">폴리곤 원 찾기</h1>
+      
+      <div className="mb-4 p-4 bg-white rounded-lg shadow">
+        <h2 className="text-lg font-semibold mb-2">설명:</h2>
+        <ul className="text-sm text-gray-700 space-y-1">
+          <li>• <span className="text-blue-600 font-semibold">파란색 원</span>: 폴리곤을 포함하는 최소 외접원</li>
+          <li>• <span className="text-green-600 font-semibold">초록색 원</span>: 외접원 내부에서 폴리곤 외부에 들어가는 최대 원</li>
+          <li>• 폴리곤의 꼭짓점을 드래그하여 모양을 변경할 수 있습니다</li>
+        </ul>
+      </div>
+      
+      <div className="bg-white rounded-lg shadow-lg p-4 mb-4">
+        <canvas
+          ref={canvasRef}
+          width={600}
+          height={500}
+          className="border border-gray-300 rounded cursor-move"
+          onMouseDown={handleMouseDown}
+          onMouseMove={handleMouseMove}
+          onMouseUp={handleMouseUp}
+          onMouseLeave={handleMouseUp}
+        />
+      </div>
+      
+      <div className="flex gap-4">
+        <button
+          onClick={resetPolygon}
+          className="px-6 py-2 bg-purple-600 text-white rounded-lg hover:bg-purple-700 transition"
+        >
+          초기화
+        </button>
+        
+        {minCircle && (
+          <div className="flex-1 p-3 bg-white rounded-lg shadow text-sm">
+            <div className="mb-1">
+              <span className="font-semibold">외접원:</span> 중심({minCircle.x.toFixed(1)}, {minCircle.y.toFixed(1)}), 반지름 {minCircle.r.toFixed(1)}
+            </div>
+            {maxInnerCircle && (
+              <div>
+                <span className="font-semibold">최대원:</span> 중심({maxInnerCircle.x.toFixed(1)}, {maxInnerCircle.y.toFixed(1)}), 반지름 {maxInnerCircle.r.toFixed(1)}
+              </div>
+            )}
+          </div>
+        )}
+      </div>
+    </div>
+  );
+};
+
+export default PolygonCircleFinder;