maxded · Nov 9, 2026 · Nov 9, 2026 · Nov 9, 2026
diff --git a/.gitignore b/.gitignore
@@ -1,8 +1,8 @@
 __pycache__/
 .venv/
 .pytest_cache/
-exports/*
-!exports/README.txt
+results/*
+!results/README.txt
 dist/
 VERSION
 requirements.txt
diff --git a/README.md b/README.md
@@ -2,7 +2,7 @@
 
 # 📊 Evolve Charts
 
-**Turn Evolve deep analysis exports into publication-ready charts**
+**Turn Evolve benchmark runs into publication-ready charts**
 
 [![Banner](./docs/images/banner.png)](https://traverseresearch.nl)
 
@@ -29,29 +29,28 @@ your web browser, and **everything stays on your computer**. Nothing is uploaded
    the libraries the tool uses (about 150 MB). This needs an internet connection, takes a few minutes and
    only happens once. Keep the window that opens running while you use the tool; close it when you're
    done.
-3. **Export your results from Evolve.** After a benchmark run, click **Export** on the score screen. Do
-   this on every GPU or setting you want to compare, using the same benchmark on each. Scores and
-   per-frame data are always included; deep analysis (per-pass timings) needs an Evolve Professional
-   licence.
-4. **Add your results.** Your browser opens Evolve Charts. Drag the export files onto *Add Evolve
-   exports*, or put them in the `exports` folder so they load automatically every time.
+3. **Run your benchmarks in Evolve.** Evolve saves every run as a `.evolve` file, in
+   `Documents\Evolve\results` on Windows. Run on every GPU or setting you want to compare, using the
+   same benchmark on each.
+4. **Add your results.** Your browser opens Evolve Charts. Drag the `.evolve` files onto *Add Evolve
+   runs*, or put them in the `results` folder so they load automatically every time.
 5. **Name your results** on their cards (for example "RTX 4070 Ti" or "Gen 1"). These names appear in
    every chart.
 
-To update, download the zip again and replace the old folder (keep your `exports` folder if you use it).
+To update, download the zip again and replace the old folder (keep your `results` folder if you use it).
 
 ## 🗺️ What's in the tool
 
-The tool reads every Evolve export: the **Export** zip from the score screen, and the files Evolve's
-command line writes (`evolve_results.csv`, `scores.csv`, the per-frame CSV and `deep_analysis.json`).
-It has a page per question, in the bar at the top:
+The tool reads the `.evolve` file Evolve saves for every run: the measurements of every frame and pass,
+the GPU sensor readings and the system information. It calculates the scores from those the way Evolve
+does. It has a page per question, in the bar at the top:
 
 | Page | What it shows | Needs |
 |---|---|---|
-| **Results** | Load and name exports; GPU, driver, OS, Evolve version and what each file contains | |
+| **Results** | Load and name runs; GPU, driver, OS, Evolve version, benchmark and what each file contains | |
 | **Scores** | Evolve's scores side by side, with the workload category each one scores, and the differences | scores |
-| **GPU time** | Where the GPU spends its time each frame, per workload category; then one workload category over the run and the passes it's made of | per-frame data; the passes need deep analysis |
-| **Passes** | Every render pass in one searchable table; tick passes to plot them over the run | deep analysis |
+| **GPU time** | Where the GPU spends its time each frame, per workload category; then one workload category over the run and the passes it's made of | per-frame data; the passes need pass timings |
+| **Passes** | Every render pass in one searchable table; tick passes to plot them over the run | pass timings |
 | **GPU sensors** | Clocks, power, temperature and more during the run, to spot throttling | per-frame data |
 | **Export** | CSV files for Excel, Google Sheets, Datawrapper and Flourish | |
 
@@ -78,13 +77,13 @@ drivers and Evolve version. **Help** explains how the numbers are made.
   because a GPU can idle between passes or run passes at the same time.
 - **Whiskers** on bar charts show the range covering 90% of frames (5th to 95th percentile).
 - Only compare results from the **same Evolve version and benchmark settings**. The tool warns you
-  when the Evolve versions or graphics APIs differ.
+  when the benchmarks, Evolve versions or graphics APIs differ.
 
 ## ❓ Troubleshooting
 
-- **A result is missing from *Passes***: its export has no deep analysis, which
-  needs an Evolve Professional licence. Its card shows ✗ Deep analysis; scores and workload
-  categories still work.
+- **A result is missing from *Passes***: its run has no pass timings. Its card shows ✗ Passes; scores
+  and workload categories still work.
+- **"was saved by a newer Evolve"**: download the latest Evolve Charts.
 - **The browser doesn't open**: open <http://localhost:8501> yourself while the black window is open.
 - **Anything else**: send a screenshot of the black window to Traverse Research.
 
@@ -93,16 +92,18 @@ drivers and Evolve version. **Help** explains how the numbers are made.
 - Run without the launcher: `uv run evolve-charts.py`. Dependencies are declared inline
   ([PEP 723](https://peps.python.org/pep-0723/)), so no virtual environment is needed.
 - Code lives in `evolve_charts/`:
-  - `loader.py` reads export zips into tidy tables.
+  - `loader.py` reads `.evolve` files (a zip of Parquet tables and a JSON document) into tidy tables.
+  - `scores.py` calculates Evolve's scores from those tables.
   - `categories.py` describes Evolve's workload categories and which pass belongs to which.
   - `stats.py` computes the summaries.
   - `charts.py` builds the Plotly figures.
   - `ui.py` is the Streamlit page.
 - The look follows Evolve's `crates/evolve/src/ui/styling.rs`: theme colours and the Spartan font are set in
   `.streamlit/config.toml`, and the fonts, logo and icon live in `static/`. Chart colours are in `charts.py`.
-- Which pass belongs to which workload category comes from Evolve's Rust source. After pass names
-  change in Evolve, regenerate the list with `python tools/sync_categories.py` (it reads
-  `../evolve/crates/evolve-frame-bucketing/src/lib.rs` by default).
+- Which pass belongs to which workload category, and the score multipliers, come from Evolve's Rust
+  source. After pass names or score calibrations change in Evolve, regenerate them with
+  `python tools/sync_categories.py` and `python tools/sync_scores.py` (both read `../evolve` by default),
+  then `uvx ruff format evolve_charts`.
 - Dependencies are locked in `evolve-charts.py.lock`; refresh it with `uv lock --script evolve-charts.py --upgrade`.
 - Tests: `uv export --script evolve-charts.py --no-hashes > requirements.txt`, then
   `uv venv && uv pip install -r requirements.txt pytest && uv run --no-project pytest`. CI runs the same.

diff --git a/docs/Android.md b/docs/Android.md
@@ -1,21 +1,17 @@
-## Automating `evolve` benchmark run on Android
+## Automating `evolve` benchmark runs on Android
 
 1. Make sure the app is installed.
-2. Use relative paths in `--export*` parameters, so that `evolve` has access to write them relative to `/sdcard/Android/data/nl.traverse_research.evolve/files/`.  Subfolders are supported, these will automatically created.
-   - It is recommended to use subfolders with the current date and time, as `evolve` will refuse to start and overwrite any export files that aleady exist.
-3. Use the `python ./scripts/run_on_android.py` script in this folder to start Evolve, like a desktop run. Common parameters to export all data from a `run-custom` are:
+2. Use the `python ./scripts/run_on_android.py` script in this folder to start Evolve with command-line
+   arguments, like a desktop run:
    ```sh
-   ts=$(date '+%F-%T')
-   python ./scripts/run_on_android.py run-custom --export-scores $ts/scores.csv --export-per-frame $ts/frames.csv --export-deep-analysis $ts/deep-analysis.json
+   python ./scripts/run_on_android.py run-custom --benchmark ray-tracing-inline
    ```
-4. When the benchmark finishes, run the following commands to pull the resulting files to your desktop:
-  ```sh
-   adb pull /sdcard/Android/data/nl.traverse_research.evolve/files/$ts/deep-analysis.json
-   adb pull /sdcard/Android/data/nl.traverse_research.evolve/files/$ts/scores.csv
-   adb pull /sdcard/Android/data/nl.traverse_research.evolve/files/$ts/frames.csv
-   ```
-   Alternatively, pull the entire timestamped folder:
-  ```sh
-   adb pull /sdcard/Android/data/nl.traverse_research.evolve/files/$ts
+3. Evolve saves every run as a `.evolve` file in
+   `/sdcard/Android/data/nl.traverse_research.evolve/files/Documents/results/`. When the benchmark finishes,
+   pull them to your desktop:
+   ```sh
+   adb pull /sdcard/Android/data/nl.traverse_research.evolve/files/Documents/results evolve-results
    ```
-   These paths can also be found directly on the phone by navigating its internal storage using MTP (when `USB for file transfer` is enabled via the notification on the phone).
+   The folder can also be found on the phone by browsing its internal storage over MTP (when
+   `USB for file transfer` is enabled via the notification on the phone).
+4. Load the `.evolve` files into Evolve Charts.
diff --git a/evolve-charts.py b/evolve-charts.py
@@ -4,9 +4,10 @@
 #     "streamlit>=1.65",
 #     "plotly>=6",
 #     "pandas>=2.2",
+#     "pyarrow>=18",
 # ]
 # ///
-"""Evolve Charts: turn Evolve deep analysis exports into charts.
+"""Evolve Charts: turn Evolve benchmark runs (.evolve files) into charts.
 
 Start it by double-clicking `Start Evolve Charts.bat` (Windows) or `start-evolve-charts.sh`
 (macOS/Linux), or run `uv run evolve-charts.py`. A page opens in your browser; nothing leaves

diff --git a/evolve-charts.py.lock b/evolve-charts.py.lock
@@ -6,17 +6,18 @@ resolution-markers = [
     "python_full_version >= '3.14' and sys_platform == 'emscripten'",
     "python_full_version >= '3.14' and sys_platform != 'emscripten' and sys_platform != 'win32'",
     "python_full_version >= '3.12' and python_full_version < '3.14' and sys_platform == 'win32'",
-    "python_full_version < '3.12' and sys_platform == 'win32'",
     "python_full_version >= '3.12' and python_full_version < '3.14' and sys_platform == 'emscripten'",
-    "python_full_version < '3.12' and sys_platform == 'emscripten'",
     "python_full_version >= '3.12' and python_full_version < '3.14' and sys_platform != 'emscripten' and sys_platform != 'win32'",
+    "python_full_version < '3.12' and sys_platform == 'win32'",
+    "python_full_version < '3.12' and sys_platform == 'emscripten'",
     "python_full_version < '3.12' and sys_platform != 'emscripten' and sys_platform != 'win32'",
 ]
 
 [manifest]
 requirements = [
     { name = "pandas", specifier = ">=2.2" },
     { name = "plotly", specifier = ">=6" },
+    { name = "pyarrow", specifier = ">=18" },
     { name = "streamlit", specifier = ">=1.65" },
 ]
 

diff --git a/evolve_charts/categories.py b/evolve_charts/categories.py
@@ -23,6 +23,11 @@
 WORKLOAD_ORDER = list(WORKLOAD_DESCRIPTIONS)
 
 
+def workload_variants(scope: str) -> list[str]:
+    """The `WorkloadCategory` variants the pass counts towards, as named in Evolve's source."""
+    return WORKLOADS.get(scope, [])
+
+
 def workloads(scope: str) -> list[str]:
     """The workload categories the pass is part of, in Evolve's order; empty if Evolve leaves it uncategorized.