{
"cells": [
{
"cell_type": "markdown",
"id": "098625d0",
"metadata": {},
"source": [
"# GAWD Language Usage\n",
"\n",
"[](https://colab.research.google.com/github/SwareUG/gh-aw-dataset-builder/blob/main/data_analysis/language_usage.ipynb)\n",
"\n",
"Language usage summary for **A Dataset of GitHub Agentic Workflow Histories: Early Adopters**.\n",
"\n",
"This follows the compact structure of the AIDev `language_usage.ipynb`: load repository and event data, compute language counts/percentages, and produce language-focused figures.\n"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "752a0d71",
"metadata": {},
"outputs": [],
"source": [
"from __future__ import annotations\n",
"\n",
"import importlib.util\n",
"import subprocess\n",
"import sys\n",
"from io import BytesIO\n",
"from pathlib import Path\n",
"from urllib.request import urlopen\n",
"\n",
"REQUIRED_PACKAGES = {\n",
" \"pandas\": \"pandas\",\n",
" \"pyarrow\": \"pyarrow\",\n",
" \"matplotlib\": \"matplotlib\",\n",
" \"seaborn\": \"seaborn\",\n",
"}\n",
"\n",
"if \"google.colab\" in sys.modules:\n",
" missing = [\n",
" package\n",
" for package, import_name in REQUIRED_PACKAGES.items()\n",
" if importlib.util.find_spec(import_name) is None\n",
" ]\n",
" if missing:\n",
" subprocess.check_call([sys.executable, \"-m\", \"pip\", \"install\", \"-q\", *missing])\n",
"\n",
"import matplotlib as mpl\n",
"import matplotlib.pyplot as plt\n",
"import pandas as pd\n",
"import seaborn as sns\n",
"from IPython.display import Image, display\n",
"\n",
"DATASET_ID = \"pavtch/gawd\"\n",
"DATASET_REVISION = \"main\"\n",
"LOCAL_RELEASE_CANDIDATES = [\n",
" Path(\"..\"),\n",
" Path(\".\"),\n",
" Path(\"../releases/gh-aw-early-adopters-source-history-2026-06-26\"),\n",
" Path(\"releases/gh-aw-early-adopters-source-history-2026-06-26\"),\n",
" Path(\"../releases/gawd\"),\n",
" Path(\"releases/gawd\"),\n",
"]\n",
"LOCAL_RELEASE_MIRROR = next(\n",
" (\n",
" path.resolve()\n",
" for path in LOCAL_RELEASE_CANDIDATES\n",
" if (path / \"data\").is_dir() and any((path / \"data\").glob(\"*.parquet\"))\n",
" ),\n",
" None,\n",
")\n",
"USE_LOCAL_MIRROR = \"google.colab\" not in sys.modules and LOCAL_RELEASE_MIRROR is not None\n",
"\n",
"\n",
"def remote_parquet_url(table_name: str) -> str:\n",
" return (\n",
" f\"https://huggingface.co/datasets/{DATASET_ID}/resolve/\"\n",
" f\"{DATASET_REVISION}/data/{table_name}.parquet\"\n",
" )\n",
"\n",
"\n",
"def local_parquet_path(table_name: str) -> Path:\n",
" if LOCAL_RELEASE_MIRROR is None:\n",
" raise FileNotFoundError(\"No local release mirror with a data/ directory was found.\")\n",
" return LOCAL_RELEASE_MIRROR / \"data\" / f\"{table_name}.parquet\"\n",
"\n",
"\n",
"def parquet_path(table_name: str) -> str:\n",
" if USE_LOCAL_MIRROR:\n",
" return str(local_parquet_path(table_name))\n",
" return remote_parquet_url(table_name)\n",
"\n",
"\n",
"def read_parquet_table(table_name: str) -> pd.DataFrame:\n",
" if USE_LOCAL_MIRROR:\n",
" return pd.read_parquet(local_parquet_path(table_name))\n",
"\n",
" try:\n",
" with urlopen(remote_parquet_url(table_name)) as response:\n",
" return pd.read_parquet(BytesIO(response.read()))\n",
" except Exception:\n",
" if LOCAL_RELEASE_MIRROR is not None:\n",
" return pd.read_parquet(local_parquet_path(table_name))\n",
" raise\n",
"\n",
"mpl.rcParams.update(\n",
" {\n",
" \"font.family\": \"serif\",\n",
" \"font.serif\": [\"Times New Roman\", \"Times\", \"DejaVu Serif\"],\n",
" \"mathtext.fontset\": \"stix\",\n",
" \"axes.linewidth\": 1.0,\n",
" \"axes.labelsize\": 14,\n",
" \"axes.titlesize\": 15,\n",
" \"xtick.major.size\": 4,\n",
" \"ytick.major.size\": 4,\n",
" \"xtick.labelsize\": 12,\n",
" \"ytick.labelsize\": 12,\n",
" \"legend.fontsize\": 11,\n",
" \"legend.title_fontsize\": 11,\n",
" \"figure.dpi\": 160,\n",
" }\n",
")\n",
"sns.set_style(\"whitegrid\")\n",
"\n",
"NOTEBOOK_DIR = Path(\"data_analysis\") if Path(\"data_analysis\").is_dir() else Path(\".\")\n",
"FIG_DIR = NOTEBOOK_DIR / \"figs\"\n",
"FIG_DIR.mkdir(parents=True, exist_ok=True)\n",
"\n",
"TOP_N = 10\n",
"\n",
"import matplotlib.dates as mdates\n",
"repository_df = read_parquet_table(\"repository\")\n",
"source_snapshot_df = read_parquet_table(\"source_markdown_file_snapshot\")\n",
"source_version_df = read_parquet_table(\"source_markdown_file_version\")\n",
"\n",
"def save_and_show(fig: plt.Figure, out: Path) -> None:\n",
" png_out = out.with_suffix(\".png\")\n",
" pdf_out = out.with_suffix(\".pdf\")\n",
" fig.savefig(png_out, bbox_inches=\"tight\", dpi=300)\n",
" fig.savefig(pdf_out, bbox_inches=\"tight\")\n",
" plt.close(fig)\n",
" display(Image(filename=str(png_out)))\n",
" print(\"Wrote\", png_out)\n",
" print(\"Wrote\", pdf_out)\n"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "dd2677bb",
"metadata": {},
"outputs": [],
"source": [
"# %%\n",
"def build_adoption_table() -> pd.DataFrame:\n",
" events = (\n",
" source_version_df.merge(\n",
" source_snapshot_df,\n",
" on=\"source_markdown_file_snapshot_id\",\n",
" how=\"inner\",\n",
" validate=\"many_to_one\",\n",
" )\n",
" .merge(repository_df, on=\"repository_id\", how=\"left\", validate=\"many_to_one\")\n",
" )\n",
" events[\"committed_at\"] = pd.to_datetime(events[\"committed_at\"], utc=True, errors=\"coerce\")\n",
" events = events.dropna(subset=[\"committed_at\"]).sort_values(\n",
" [\"repository_id\", \"committed_at\", \"rank\"]\n",
" )\n",
" adoption = (\n",
" events.groupby(\"repository_id\", as_index=False)\n",
" .agg(\n",
" repo_full_name=(\"repo_full_name\", \"first\"),\n",
" main_language=(\"main_language\", \"first\"),\n",
" stars=(\"stars\", \"first\"),\n",
" forks=(\"forks\", \"first\"),\n",
" adoption_date=(\"committed_at\", \"first\"),\n",
" observed_versions=(\"source_markdown_file_version_id\", \"count\"),\n",
" )\n",
" .sort_values(\"adoption_date\")\n",
" )\n",
" adoption[\"language\"] = adoption[\"main_language\"].fillna(\"Unknown\")\n",
" adoption[\"adoption_month\"] = (\n",
" adoption[\"adoption_date\"].dt.tz_convert(None).dt.to_period(\"M\").dt.to_timestamp()\n",
" )\n",
" return adoption\n",
"\n",
"\n",
"def language_counts(adoption: pd.DataFrame) -> pd.DataFrame:\n",
" counts = adoption[\"language\"].value_counts(dropna=False).reset_index()\n",
" counts.columns = [\"language\", \"repositories\"]\n",
" counts[\"percentage\"] = counts[\"repositories\"] / counts[\"repositories\"].sum() * 100\n",
" return counts\n",
"\n",
"\n",
"def plot_total_language_percentages(counts: pd.DataFrame, top_n: int = TOP_N) -> None:\n",
" plot_df = counts.head(top_n).copy()\n",
" fig, ax = plt.subplots(figsize=(7, 0.45 * len(plot_df) + 1.4))\n",
" sns.barplot(\n",
" data=plot_df,\n",
" y=\"language\",\n",
" x=\"percentage\",\n",
" orient=\"h\",\n",
" color=\"#1C3F60\",\n",
" ax=ax,\n",
" )\n",
" ax.set_xlabel(\"Percentage of Repositories (%)\")\n",
" ax.set_ylabel(None)\n",
" ax.set_title(\"Top Languages Used by Observed GH-AW Adopters\", weight=\"bold\")\n",
" ax.set_xlim(0, plot_df[\"percentage\"].max() * 1.15)\n",
" ax.grid(False)\n",
" sns.despine(ax=ax)\n",
" fig.tight_layout()\n",
" out = FIG_DIR / \"gawd_total_language_percentages_top.png\"\n",
" save_and_show(fig, out)\n",
"\n",
"\n",
"def plot_language_cumulative(adoption: pd.DataFrame, top_n: int = 8) -> None:\n",
" top_languages = adoption[\"language\"].value_counts().head(top_n).index\n",
" plot_df = adoption.copy()\n",
" plot_df[\"language_group\"] = plot_df[\"language\"].where(\n",
" plot_df[\"language\"].isin(top_languages), \"Other\"\n",
" )\n",
" monthly = (\n",
" plot_df.groupby([\"language_group\", \"adoption_month\"], as_index=False)\n",
" .agg(new_repositories=(\"repository_id\", \"count\"))\n",
" .sort_values([\"language_group\", \"adoption_month\"])\n",
" )\n",
" all_months = pd.date_range(\n",
" adoption[\"adoption_month\"].min(), adoption[\"adoption_month\"].max(), freq=\"MS\"\n",
" )\n",
" complete_index = pd.MultiIndex.from_product(\n",
" [sorted(monthly[\"language_group\"].unique()), all_months],\n",
" names=[\"language_group\", \"adoption_month\"],\n",
" )\n",
" monthly = (\n",
" monthly.set_index([\"language_group\", \"adoption_month\"])\n",
" .reindex(complete_index, fill_value=0)\n",
" .reset_index()\n",
" )\n",
" monthly[\"cumulative_repositories\"] = monthly.groupby(\"language_group\")[\n",
" \"new_repositories\"\n",
" ].cumsum()\n",
"\n",
" fig, ax = plt.subplots(figsize=(12, 7))\n",
" sns.lineplot(\n",
" data=monthly,\n",
" x=\"adoption_month\",\n",
" y=\"cumulative_repositories\",\n",
" hue=\"language_group\",\n",
" marker=\"o\",\n",
" linewidth=2.2,\n",
" palette=\"tab10\",\n",
" ax=ax,\n",
" )\n",
" ax.set_title(\"Cumulative Observed GH-AW Adoption by Main Language\", weight=\"bold\")\n",
" ax.set_xlabel(\"First observed adoption month\")\n",
" ax.set_ylabel(\"Cumulative repositories\")\n",
" ax.xaxis.set_major_locator(mdates.AutoDateLocator())\n",
" ax.xaxis.set_major_formatter(mdates.ConciseDateFormatter(ax.xaxis.get_major_locator()))\n",
" ax.legend(title=\"Main language\", bbox_to_anchor=(1.02, 1), loc=\"upper left\")\n",
" sns.despine(ax=ax)\n",
" fig.tight_layout()\n",
" out = FIG_DIR / \"gawd_cumulative_language_usage.png\"\n",
" save_and_show(fig, out)\n"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "016cf9c4",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" language | \n",
" repositories | \n",
" percentage | \n",
"
\n",
" \n",
" \n",
" \n",
" | 0 | \n",
" TypeScript | \n",
" 52 | \n",
" 19.847328 | \n",
"
\n",
" \n",
" | 1 | \n",
" Go | \n",
" 41 | \n",
" 15.648855 | \n",
"
\n",
" \n",
" | 2 | \n",
" C# | \n",
" 32 | \n",
" 12.213740 | \n",
"
\n",
" \n",
" | 3 | \n",
" Python | \n",
" 29 | \n",
" 11.068702 | \n",
"
\n",
" \n",
" | 4 | \n",
" Makefile | \n",
" 21 | \n",
" 8.015267 | \n",
"
\n",
" \n",
" | 5 | \n",
" F# | \n",
" 18 | \n",
" 6.870229 | \n",
"
\n",
" \n",
" | 6 | \n",
" Rust | \n",
" 13 | \n",
" 4.961832 | \n",
"
\n",
" \n",
" | 7 | \n",
" JavaScript | \n",
" 11 | \n",
" 4.198473 | \n",
"
\n",
" \n",
" | 8 | \n",
" Java | \n",
" 10 | \n",
" 3.816794 | \n",
"
\n",
" \n",
" | 9 | \n",
" C++ | \n",
" 7 | \n",
" 2.671756 | \n",
"
\n",
" \n",
" | 10 | \n",
" Shell | \n",
" 5 | \n",
" 1.908397 | \n",
"
\n",
" \n",
" | 11 | \n",
" Kotlin | \n",
" 4 | \n",
" 1.526718 | \n",
"
\n",
" \n",
" | 12 | \n",
" C | \n",
" 3 | \n",
" 1.145038 | \n",
"
\n",
" \n",
" | 13 | \n",
" HTML | \n",
" 2 | \n",
" 0.763359 | \n",
"
\n",
" \n",
" | 14 | \n",
" Vue | \n",
" 2 | \n",
" 0.763359 | \n",
"
\n",
" \n",
" | 15 | \n",
" Ruby | \n",
" 2 | \n",
" 0.763359 | \n",
"
\n",
" \n",
" | 16 | \n",
" Elixir | \n",
" 1 | \n",
" 0.381679 | \n",
"
\n",
" \n",
" | 17 | \n",
" Dart | \n",
" 1 | \n",
" 0.381679 | \n",
"
\n",
" \n",
" | 18 | \n",
" Erlang | \n",
" 1 | \n",
" 0.381679 | \n",
"
\n",
" \n",
" | 19 | \n",
" Swift | \n",
" 1 | \n",
" 0.381679 | \n",
"
\n",
" \n",
" | 20 | \n",
" Jupyter Notebook | \n",
" 1 | \n",
" 0.381679 | \n",
"
\n",
" \n",
" | 21 | \n",
" PHP | \n",
" 1 | \n",
" 0.381679 | \n",
"
\n",
" \n",
" | 22 | \n",
" Lua | \n",
" 1 | \n",
" 0.381679 | \n",
"
\n",
" \n",
" | 23 | \n",
" R | \n",
" 1 | \n",
" 0.381679 | \n",
"
\n",
" \n",
" | 24 | \n",
" Julia | \n",
" 1 | \n",
" 0.381679 | \n",
"
\n",
" \n",
" | 25 | \n",
" Unknown | \n",
" 1 | \n",
" 0.381679 | \n",
"
\n",
" \n",
"
\n",
"