I was confused by a sudden lack of results in my subsequence search, and have been able to identify a specific commit that seems to have broken the following code by methodically reinstalling and going forward one commit at a time until the below test breaks:
pip uninstall dtaidistance
git clone https://github.com/wannesm/dtaidistance.git && cd dtaidistance
git checkout <next commit to test>
sed -i '' 's/double/float/g' dtaidistance/jinja/generate.py dtaidistance/lib/DTAIDistanceC/DTAIDistanceC/dd_globals.h
cd dtaidistance/jinja && make && cd ../..
pip install .
< restart Jupyter server and kernel completely >
Consider this standalone reproducible "test" case:
This is confirmed to be working at commit 26c39622a269cd2bff9ddc42683b21ae3b2533e5
But it is not working one commit later at 31f77034bd7687cbcb2820be008065bd8d1d6ff8
If you do print(len(list(best))) at this broken commit, you'll see there's only 1 item, with a distance of infinity and an idx of -1. Yet in the test data above there is exactly one series with a distance less than the maxDistance. If you change k = None to a positive integer, it does seem to work.
I was confused by a sudden lack of results in my subsequence search, and have been able to identify a specific commit that seems to have broken the following code by methodically reinstalling and going forward one commit at a time until the below test breaks:
Consider this standalone reproducible "test" case:
This is confirmed to be working at commit 26c39622a269cd2bff9ddc42683b21ae3b2533e5
But it is not working one commit later at 31f77034bd7687cbcb2820be008065bd8d1d6ff8
If you do
print(len(list(best)))
at this broken commit, you'll see there's only 1 item, with a distance of infinity and an idx of -1. Yet in the test data above there is exactly one series with a distance less than the maxDistance. If you change k = None to a positive integer, it does seem to work.