stats.yaml recomputed on the 320-case training split (all frames, all points)

data/stats.yaml

@@ -1,36 +1,39 @@
+# TurtLES normalisation statistics: train split(s), 320 cases,
+# all 400 frames and all points per case, weighting = points.
+# delta = x(t+1) - x(t) between consecutive saved frames (399 per case).
 mean:
-  p: 0.99131
-  rho: 0.99181
-  u: 0.19128
-  v: 0.00082661
-  w: -1.6723e-05
+  p: 0.99107
+  rho: 0.99157
+  u: 0.18936
+  v: 0.00094513
+  w: 6.8113e-06
 rms:
-  p: 0.99146
-  rho: 0.99193
-  u: 0.21833
-  v: 0.070125
-  w: 0.030908
+  p: 0.99122
+  rho: 0.99168
+  u: 0.21756
+  v: 0.070636
+  w: 0.031052
 std:
-  p: 0.017135
-  rho: 0.014972
-  u: 0.10525
-  v: 0.070121
-  w: 0.030908
+  p: 0.017384
+  rho: 0.015194
+  u: 0.10711
+  v: 0.07063
+  w: 0.031052
 mean_delta:
-  p: 4.1888e-07
-  rho: 1.9519e-07
-  u: -1.8495e-06
-  v: 7.6061e-07
-  w: 5.169e-07
+  p: -9.1876e-08
+  rho: -6.8633e-08
+  u: 2.6201e-07
+  v: -2.4647e-07
+  w: 2.5538e-08
 rms_delta:
-  p: 0.0010532
-  rho: 0.0011146
-  u: 0.0092878
-  v: 0.010322
-  w: 0.010015
+  p: 0.0010483
+  rho: 0.0011113
+  u: 0.0092925
+  v: 0.010306
+  w: 0.010003
 std_delta:
-  p: 0.0010532
-  rho: 0.0011146
-  u: 0.0092878
-  v: 0.010322
-  w: 0.010015
+  p: 0.0010483
+  rho: 0.0011113
+  u: 0.0092925
+  v: 0.010306
+  w: 0.010003