Oriol Nieto

Ideally, any default parameter (like [this](https://github.com/rabitt/motif/blob/master/motif/core.py#L307), [this](https://github.com/rabitt/motif/blob/master/motif/core.py#L482), or [this](https://github.com/rabitt/motif/blob/master/motif/core.py#L586)). But I understand how tedious this can become. You may know better than anyone which parameters are the most important ones...

Sounds good. If you create a new PR, could you do it in the 0.1.3-dev branch? Thanks!

Ok, this should be ready for v0.1.6 (04ff0dc078867a3bd19803861068b6ad730751ec). Thanks!

PS: your visualization tool is awesome (of course, your algorithm too!).

mmmh might be a bug? I assume you refer to this part of the code in MSAF: https://github.com/urinieto/msaf/blob/master/msaf/algorithms/scluster/main2.py#L58 Right now `scluster_k` is the fixed number of unique segment types (i.e.,...

sounds great, thanks! On Feb 17, 2018 4:33 PM, "Cheng-i Wang" wrote: > It would be this part here in the original implementation, > https://github.com/bmcfee/laplacian_segmentation/blob/ > 94a2c347d841554cf4d1a3b0f39b666f23ecf576/code/segmenter.py#L492-L514 > , where...

Thanks for reporting. I need to do a full update of MSAF, since several libraries are outdated as of now. I recommend you creating a virtual env with those old...

Thanks so much for the help, Dorien, and for the nice comment! :)

Hi, seems like your numpy version is not compatible with `msaf`. According to [this](https://stackoverflow.com/questions/20518632/importerror-numpy-core-multiarray-failed-to-import), you may be able to solve it by doing: ``` pip install numpy==1.19.3 ``` Hope this...

Thanks for reporting this. I will try to work on a workaround soon. This seems indeed problematic and should be fixed asap.