search

ahwillia / affinewarp

An implementation of piecewise linear time warping for multi-dimensional time series alignment
MIT License
162 stars 36 forks source link

Fix exponential decay in the random search #11

Closed kushbanga closed 4 years ago

kushbanga commented 4 years ago

Currently the perturbation sizes in the random search grow exponentially while the comment above the changed line suggests they should decay?

ahwillia commented 4 years ago

Thank you so much for this! I had to change the default temperature settings to make the fitting algorithm work again so I ended up fixing just now in this commit -- f733f02637ecd5f57ddf54237cfa31500bac4e3f. Go ahead and do a git pull and you should be good to go. Thanks again, and don't hesitate to reach out again if you encounter problems or have any questions.