Etienne Ackermann
Etienne Ackermann
This is now quickly becoming a high priority, since we need this for elegant decoding cross-validation. @sillyproject you can also learn about this when I finish this task.
Slice indexing is now done, so all that remains is EpochArray indexing. However, an additional point of discussion is how to handle slices that extend beyond an object's range. If...
^ this is consistent with default Python behavior though. See, e.g., http://stackoverflow.com/questions/509211/explain-pythons-slice-notation
Also look into this for `predict_proba`, and any others that may require it.
So in particular, I would like to be able to call `npl.plot(analogsignalarray, compact=True)` to plot any of the rows above.
Minimal (non-)working example: ``` import matplotlib as mpl import matplotlib.pyplot as plt import matplotlib.patches as patches import numpy as np def testplot(ax=None): if ax is None: ax = plt.gca() ax.plot(np.random.rand(10),...
This also shows a potentially nice wrapper to profile things. Could even be a built-in utility?
If I had asa.mean(axis=1), then I could use _that_ one to plot my LFP instead of picking out a particular channel, as mentioned in #126
Yes, that's what I was asking for, but of course, it might make sense to return a new AnalogSignalArray. This points to an unsolved problem, actually. How should we store...
This might call for a re-index function, to reindex time to some baseline or starting point. E.g., 0,1,2—reindex—3,4,5,6,7—reindex—8,9 --> 0,1,2—0,1,2,3,4—0,1