Repository navigation
Issue following tutorial from video #655
Description
I'm following the youtube tutorial of syndy and I believe you need to update the tutorial code ( https://github.com/dynamicslab/pysindy/blob/master/examples/15_pysindy_lectures.ipynb ) because it seems feature_names isn't a parameter of SINDy class anymore. At least I had an error
TypeError Traceback (most recent call last)
Cell In[10], line 3
1 feature_names = ['x','y', 'z']
2 opt = ps.STLSQ(threshold=0)
----> 3 model = ps.SINDy(feature_names = feature_names, optimizer=opt)
4 model.fit(x_train, t = dt)
5 model.print()
TypeError: SINDy.init() got an unexpected keyword argument 'feature_names'
Well, I know it's a silly thing to report, but I believe it will make the tutorial better to follow.
Reacted by Aru
Activity
nick0413 commentedon Nov 7, 2025
I am getting the same thing right now, and if the
feature_namesis no longer part of expected usage, then the docs are outdated. So I think it might still be a useful issue.mickpires commentedon Nov 8, 2025
it seems the docs is outdated regarding this part because it cites that the parameter exists.
Jacob-Stevens-Haas commentedon Nov 9, 2025
Yes, That notebook should be updated. The parameter has been moved from
SINDy.__init__toSINDy.fit, because its a data-dependent parameter.The current plan was to break that notebook into smaller pieces, since it addresses several issues and it's hard to run a long notebook in CI. However, the youtube video has been a big draw and an important piece of the documentation. Maybe we need to think how to better redirect people to the appropriate documentation.
Aru-WaveRider commentedon Jan 17, 2026
The versions shown in videos are older ones. I faced the same issue and being a newbie in using such set ups, I kept struggling with this for 2 months! They should either mention the points on the repo or add comments on the youtube videos mentioning the changes.
So much time wasted :( :( :(
Jacob-Stevens-Haas commentedon Jan 20, 2026
@Aru-WaveRider I'm sorry you spent so long struggling! Always feel free to ask for help here or in the discussions page! Another pro tip is that, regardless of documentation online,
help(obj)in interactive python will always bring up the documentation attached to the particular object you're using. So if you try:gives the docs on how to instantiate a model, and
gives the docs on how to fit a model for the version you are using.