Maarten Breddels

Opened https://github.com/jupyter-widgets/ipywidgets/pull/2338

Normalized would be nice, since there may be a scale associated to the image that does not correspond to pixels. And you never know the try image size in the...

With parquet is slightly faster, but I cannot run the filtered part 

ok, i pushed a failing test, the situation where it failed is.. a filtered dataframe, where 1 of the tasks does pre-filtering (like sum), and the other does not (nop...

Indeed, it's being converted on the fly when doing many operations. How did it confused you @NickCrews ?

I have to say, after going through the pandas documentation, and trying it out in the notebook, I don't understand what this should do. @JovanVeljanoski maybe you understand?

yeah, don't worry, that's unrelated thank you for your first contribution!

Hi, I would like to avoid that pattern as much as possible because it is not that efficient by design. It's either going to lead to random access to the...

It would help us a lot if also you could specify the specific use case. We have some ideas on how to implement it efficiently, but the more examples/use cases...

I'm looking at https://github.com/vaexio/vaex/blob/192b812ac1c9fbdefbb89a987d93b35ca83a5729/packages/vaex-core/vaex/dataset_mmap.py#L90 and you may be right. Could you try: ``` for f in df.mapping_map.values(): f.close() ```