Continuous benchmarking

.github/workflows/benchmark.yml

@@ -0,0 +1,62 @@
+name: benchmark
+on:
+  push:
+    branches:
+      - master
+  pull_request:
+    branches:
+      - master
+
+jobs:
+  do-benchmark:
+    name: Benchmark runner
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout folder to get test files
+        uses: actions/checkout@v3
+        if: github.event_name == 'pull_request'
+        with:
+          fetch-depth: 0
+          ref: ${{ github.event.pull_request.head.ref }}
+
+      - uses: actions/checkout@v3
+        if: github.event_name == 'push'
+        with:
+          fetch-depth: 0
+
+      - name: Rename checked out topojson package
+        run: mv topojson do_later_topojson
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: "3.10"
+      - name: Install test dependencies
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install geopandas shapely numpy fire altair altair_saver okab
+
+      - name: Install latest topojson version from pypi
+        run: python -m pip install topojson
+      - name: Benchmark last released version!
+        run: python tests/benchmark_compute.py --version=last-released
+
+      - name: Install master repo topojson from git
+        run: python -m pip install git+https://github.com/mattijn/topojson.git --upgrade
+      - name: Benchmark version in github master!
+        run: python tests/benchmark_compute.py --version=master
+
+      - name: Rename moved topojson package
+        run: mv do_later_topojson topojson
+      - name: Benchmark this PR!
+        run: python tests/benchmark_compute.py --version=PR
+
+      - name: Create visz from benchmark results!
+        run: python tests/benchmark_visz.py
+
+      - name: Commit benchmark png
+        uses: EndBug/add-and-commit@v9
+        with:
+          default_author: github_actions
+          message: "generate benchmark chart"
+          add: "tests/benchmark_chart.png --force"

.github/workflows/test.yml

@@ -2,11 +2,11 @@ name: test
 on:
   push:
     branches:
-    - master
+      - master
   pull_request:
     branches:
-    - master
-    
+      - master
+
 jobs:
   do-explore:
     name: Feedback runner
@@ -23,7 +23,7 @@ jobs:
         run: |
           ls ${{ github.workspace }}
       - run: echo "🍏 This job's status is ${{ job.status }}."
-      
+
   do-test:
     name: Test on Python
     runs-on: ubuntu-latest
@@ -41,11 +41,11 @@ jobs:
       - name: Install test dependencies
         run: |
           python -m pip install --upgrade pip
-          python -m pip install flit codecov pytest flake8     
+          python -m pip install flit codecov pytest flake8
       - name: Install package dependencies
-        run: flit install --deps develop        
+        run: flit install --deps develop
       - name: Overrule shapely version from matrix
-        run: python -m pip install "shapely==${{ matrix.shapely-version }}"          
+        run: python -m pip install "shapely==${{ matrix.shapely-version }}"
       - name: Lint with flake8
         run: flake8 topojson
         continue-on-error: true

tests/benchmark_compute.py

@@ -0,0 +1,52 @@
+import fire
+import json
+from time import process_time
+from pathlib import Path
+import geopandas
+import topojson
+
+
+def time_topology(data):
+    t_start = process_time()
+    _ = topojson.Topology(data)
+    t_stop = process_time()
+    return t_stop - t_start
+
+
+def time_version(version):
+    """
+    The following list contain files are being timed. The files filed should be able to
+    open by geopandas.read_file(). The characteristic of each file is as follow:
+    - nybb: a few detailed long linestrings
+    - naturalearth_lowres: similar alike linestrings, not too long
+    - mesh2d: many very short linestrings
+    """
+    files_to_time = [
+        geopandas.datasets.get_path("nybb"),
+        geopandas.datasets.get_path("naturalearth_lowres"),
+        "tests/files_geojson/mesh2d.geojson",
+    ]
+
+    # apply 3x timing to each file and collect result in list
+    list_times = []
+    for file_to_time in files_to_time:
+        gdf_data = geopandas.read_file(file_to_time)
+        for _ in range(3):
+            time_of_file = time_topology(gdf_data)
+            list_times.append(
+                {
+                    "file": Path(file_to_time).stem,
+                    "time": time_of_file,
+                    "version": version,
+                }
+            )
+
+    # save timings within test folder, these files will not be kept, but only used to
+    # create a visualization in benchmark_visz.py
+    with open(f"tests/timings_{version}.json", "w") as f:
+        json.dump(list_times, f)
+
+
+if __name__ == "__main__":
+
+    fire.Fire(time_version)

tests/benchmark_visz.py

@@ -0,0 +1,42 @@
+import fire
+from pathlib import Path
+import altair as alt
+import pandas as pd
+from okab.saver import OkabSaver
+
+
+def stats_to_visz():
+    # the following 3 versions are compared, namely the last released version on pypi
+    # the version that is currently on master in github and the version in the PR being
+    # processed.
+    versions = ["last-released", "master", "PR"]
+
+    # collect all timings and create a pandas dataframe
+    list_df = []
+    for version in versions:
+        fp_version = Path(f"tests/timings_{version}.json").resolve()
+        fp_version = fp_version.as_posix()
+        df_version = pd.read_json(fp_version, orient="records")
+        list_df.append(df_version)
+    df = pd.concat(list_df)
+
+    # parse to altair chart
+    chart = (
+        alt.Chart(df, title="timing of files (s) across version", height=150)
+        .mark_circle()
+        .encode(
+            x=alt.X("version:N", title=None),
+            y=alt.Y("time:Q", title=None, scale=alt.Scale(type="log")),
+            color="version:N",
+            column=alt.Column("file:N", title=None),
+        )
+        .resolve_scale(x="independent")
+    )
+
+    # save chart as png using okab as save method
+    chart.save("tests/benchmark_chart.png", method=OkabSaver, scale_factor=2)
+
+
+if __name__ == "__main__":
+
+    fire.Fire(stats_to_visz)