Create DATA_DICTIONARY_EXCERPT.md
Browse files- DATA_DICTIONARY_EXCERPT.md +52 -0
DATA_DICTIONARY_EXCERPT.md
ADDED
|
@@ -0,0 +1,52 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
EX-NIHILO-CODE-001 — Public Sample Dictionary
|
| 2 |
+
Sample grain: one public GitHub repository per row
|
| 3 |
+
Sample size: 100 of 339 full-index rows
|
| 4 |
+
Order: stars descending
|
| 5 |
+
Composition: 97 verified-cut rows; 3 below the 100-merged-PR threshold
|
| 6 |
+
Data pull: 2026-09-16
|
| 7 |
+
|
| 8 |
+
This is a public dictionary excerpt for the Kaggle sample, not the complete paid documentation.
|
| 9 |
+
|
| 10 |
+
Core fields
|
| 11 |
+
Field Meaning
|
| 12 |
+
repo_id Stable GitHub repository identifier
|
| 13 |
+
owner Repository owner or organization
|
| 14 |
+
repo_name Repository name
|
| 15 |
+
source_url Canonical public GitHub URL
|
| 16 |
+
language Named language used for the language-share gate
|
| 17 |
+
stars GitHub star count at pull time
|
| 18 |
+
forks GitHub fork count at pull time
|
| 19 |
+
license_spdx SPDX identifier returned by GitHub
|
| 20 |
+
commercial_use_ok Derived allowlist flag; not a legal opinion or relicense
|
| 21 |
+
size_kb Git disk usage reported by GitHub; not LOC
|
| 22 |
+
estimated_loc Linguist language bytes divided by language-specific bytes per line
|
| 23 |
+
primary_language_pct Named-language bytes divided by total linguist bytes
|
| 24 |
+
merged_pr_count GitHub merged PR count when successfully queried
|
| 25 |
+
pr_verified Whether the PR count was successfully verified
|
| 26 |
+
has_tests Test-directory signal when inspected
|
| 27 |
+
has_ci CI-file or CI-directory signal when inspected
|
| 28 |
+
last_commit_date Repository push date used for recency screening
|
| 29 |
+
data_pull_date UTC date of the index observation
|
| 30 |
+
Admission rules
|
| 31 |
+
license_spdx is MIT, Apache-2.0, BSD-2-Clause, or BSD-3-Clause
|
| 32 |
+
estimated_loc is 100,000–10,000,000
|
| 33 |
+
primary_language_pct is at least 0.40
|
| 34 |
+
Repository is current, not archived, and not a fork
|
| 35 |
+
Verified cut additionally requires merged_pr_count >= 100 and pr_verified=true
|
| 36 |
+
Premium cut additionally requires stars >= 1000 and estimated_loc >= 250000
|
| 37 |
+
LOC methodology
|
| 38 |
+
estimated_loc is calculated from GitHub linguist language bytes:
|
| 39 |
+
|
| 40 |
+
Rust: bytes ÷ 42
|
| 41 |
+
Go: bytes ÷ 38
|
| 42 |
+
C++: bytes ÷ 48
|
| 43 |
+
Java: bytes ÷ 52
|
| 44 |
+
COBOL: bytes ÷ 72
|
| 45 |
+
MATLAB: bytes ÷ 44
|
| 46 |
+
size_kb is Git disk usage and can include repository history and objects. It must not be multiplied by 30 or represented as source line count. See size_kb_vs_estimated_loc.svg.
|
| 47 |
+
|
| 48 |
+
Important interpretation notes
|
| 49 |
+
A permissive SPDX identifier is an index admission signal, not a relicense or substitute for legal review.
|
| 50 |
+
Empty optional fields are allowed; do not advertise “no nulls.”
|
| 51 |
+
golang/go uses Gerrit. Its GitHub merged-PR count is not its total contribution volume.
|
| 52 |
+
This free archive contains no pull-request harvest.
|