Jacob

The EllipticEnvelope is a batch method (only implements the `fit` method for training). You can still use it but you need to create and maintain a window of data for...

The `BatchIncrementalClassifier` is an ensemble method for batch-incremental learning. Each member of the ensemble is trained on a different batch (window) of data. The method is quite simple as it...

The example in the [documentation](https://scikit-multiflow.readthedocs.io/en/stable/api/generated/skmultiflow.meta.BatchIncrementalClassifier.html#skmultiflow.meta.BatchIncrementalClassifier) shows how to call `partial_fit` and `predict`. For your example: ```python import numpy as np from skmultiflow.meta import BatchIncrementalClassifier from skmultiflow.data import DataStream from sklearn.covariance...

Travis fails due to the latest version of sklearn. I will update a temporary workaround as part of the other open PR. I will upload the actual fix later today.

Thanks for the suggestion @mertozer94 I am trying to figure out where/how we could fit this type of methods in the current structure. Any comment @jmread?

Sorry for the delay. Both approaches are interesting. The simpler option (IMO) is an extension of the stream class that would yield the data in the described manner. This might...

I vote to return `0` given that most methods will fail if they get a `NAN`. Additionally, at the implementation level, we could `shift` the start point to the second...

Thanks for the information @martindurant As you mention, data ingestion is something that we are looking forward to improving. I talked with some people at AnacondaCON regarding this topic and...

Hi @sam1357035 Please provide an example of the error log or MWE. Note: The `StreamingRandomPatchesClassifier` is designed for cases where there is a reasonably large number of features, in low...

Those are really interesting examples and thanks for the viz stack explanation, it can be confusing at first. I think `Panel` is worth considering if it allows to easily display...