viz

Visualization

This tool displays the detectors and errors from a Tesseract decoding run in 3D.

Generating the JSON

Use the --verbose flag when running the decoder and filter the output to include only the lines used by the converter script:

bazel build src:all && \
./bazel-bin/src/tesseract \
  --sample-num-shots 1 --det-order-seed 13267562 --pqlimit 10000 --beam 1 --num-det-orders 20 \
  --circuit testdata/colorcodes/r\=9\,d\=9\,p\=0.002\,noise\=si1000\,c\=superdense_color_code_X\,q\=121\,gates\=cz.stim \
  --sample-seed 717347 --threads 1 --verbose | \
  grep -E 'Error|Detector|activated_errors|activated_detectors' > logfile.txt

python viz/to_json.py logfile.txt -o logfile.json

The --no-det-order-bfs flag is compatible with visualization logs. BFS-based detector ordering is now enabled by default, so include this flag only if you want to disable it. Make sure --verbose is enabled so the detector coordinates are printed for to_json.py to parse.

The to_json.py script produces logfile.json, which contains the detector coordinates and animation frames for the viewer.

Viewing

Open viz/index.html in a modern browser. It will automatically try to load logfile.json from the same directory. If the file picker is used, any JSON produced by to_json.py can be visualized.