Fix for #20 with --det-order-bfs, and update visualizer

[noajshu]

Fix #20 by adding --det-order-bfs argument.
Apparently it can improve performance for the same LER as well:

Before change:
bazel build src:all && time ./bazel-bin/src/tesseract --pqlimit 200000 --beam 15 --beam-climbing --num-det-orders 16  --sample-num-shots 10000 --det-order-seed 13267562 --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 60  --print-stats
num_shots = 10000 num_low_confidence = 0 num_errors = 52 total_time_seconds = 64829.16896500005

After change:
bazel build src:all && time ./bazel-bin/src/tesseract --pqlimit 200000 --beam 5 --num-det-orders 20  --sample-num-shots 10000 --det-order-seed 13267562 --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 60 --det-order-bfs
num_shots = 10000 num_low_confidence = 2 num_errors = 26 total_time_seconds = 46809.56727799983

Plus a few updates for the visualizer, now it should load a default logfile automatically!

[LalehB]

LGTM!
Thanks Noah 👍🏻
btw is viz/logfile.json the file with the color code example?

[noajshu]

btw is viz/logfile.json needed?
This is needed in order for the html to be able to load a default file for visualization. It is on the smaller side so I figured it would be OK to store in the github.

[noajshu]

also tried without the dfs det order and the tighter beam settings:

time ./bazel-bin/src/tesseract --pqlimit 200000 --beam 5 --num-det-orders 20  --sample-num-shots 10000 --det-order-seed 13267562 --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 60 

num_shots = 10000 num_low_confidence = 5 num_errors = 43 total_time_seconds = 55060.10312700005

[noajshu]

BTW -- adding --no-revisit-dets like in the paper did not help the older version of det ordering:

bazel build src:all && time ./bazel-bin/src/tesseract --no-revisit-dets --pqlimit 200000 --beam 15 --beam-climbing --num-det-orders 16  --sample-num-shots 10000 --det-order-seed 13267562 --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 60  --print-stats

# run 1
num_shots = 10000 num_low_confidence = 0 num_errors = 47 total_time_seconds = 62430.72689600012
# run 2
num_shots = 10000 num_low_confidence = 0 num_errors = 52 total_time_seconds = 65866.87115100007

With the --det-order-bfs it is of course faster as reported above:

bazel build src:all && time ./bazel-bin/src/tesseract --pqlimit 200000 --beam 5 --num-det-orders 20  --sample-num-shots 10000 --det-order-seed 13267562 --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 60 --det-order-bfs

num_shots = 10000 num_low_confidence = 2 num_errors = 26 total_time_seconds = 47688.57385300003

and with the --det-order-bfs and --no-revisit-dets it is the best

bazel build src:all && time ./bazel-bin/src/tesseract --pqlimit 200000 --beam 5 --num-det-orders 20  --sample-num-shots 10000 --no-revisit-dets --det-order-seed 13267562 --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 60 --det-order-bfs

num_shots = 10000 num_low_confidence = 0 num_errors = 27 total_time_seconds = 40312.73824599993