vincent d warmerdam

If I think about the "average user" then I imagine it's far more likely that you'll be working with images on disk. You'll need to have them in memory if...

Cool. I can't promise when I'll have a look, but it's nice to know it's around 👍

Note! While still experimental, I've started work on [embetter](https://github.com/koaning/embetter). It's a library that should make it easier to just grab and toy around with some useful embeddings and it offers...

Could you share the first few rows of the `ready.csv` file? Have you added `x` and `y` columns like in the readme?

Normally, you would consider subclassing because parts can be re-used. Scikit-learn has a `ClassifierMixin` so that all classifiers are consistent for example. The thing with this package though ... it's...

We could add an optional dependency and an optional helper for datetime parsing using [datefinder](https://github.com/akoumjian/datefinder). Given that we have dates in a proper iso-standard it would be nice to have...

What I'm proposing here is certainly similar to the grouper, but I think I'm going for something that is more minimal. I'm also not 100% sure that I like their...

The doc makes it clear, but a user would need to leave the notebook and find that page in order to move on. I'd prefer to just use plain words....

My preferred attitude here is only to start supporting extra features if there is a strong need for it. I added this issue because I had a need for basic...

Now that I am thinking about it, maybe this is a more consistent API: ```python clump.mutate(rounded_date = lambda d: round_dt_str(d['dt'], by='hour')) ``` The benefits: 1. This keeps the API functional/less...