Joe Kington

If you're not in a notebook, do you know which matplotlib backend you're using? What's happening (notebook or not) is that things are being refreshed too frequently. Hover mode in...

There is no "default backend", for what it's worth. If nothing is specified, matplotlib chooses based on the OS and what packages are available. From the stacktraces, it looks like...

I still haven't been able to reproduce the issues you're seeing, unfortunately. I've gotten access to a Windows laptop and tested things there. Everything seems extremely responsive for me. However,...

It's perfectly responsive for me with exactly that code yes. Especially when zoomed in. However, note that `sns.heatmap` plots things with `pcolormesh` which means that matplotlib only considers clicks on...

To get a sense of what I'm talking about, try using `imshow`. Notice that it will trigger anywhere within the cell, not just at the edges: ``` import numpy as...

@steel3d - For what it's worth, I've considered hacking around the limitations of `pcolormesh` (or, more accurately, the `QuadMesh` artist) interactions in the past so that logical cells (i.e. what...

@jondo - The python3 related commits were merged in a separate pull request by @sapus well before this pull request (e.g. https://github.com/joferkington/mpldatacursor/commit/c0e1950e547a4e6af48fc319c7b516a930049942 ) Note that there were a couple of...

For date axes, matplotlib uses an internal representation of dates. The original `datetime` instances are converted into this floating-point-based date representation. You can convert the coordinates you're getting back into...

@farzanullah - The code in these demos works with "raw" numpy arrays. You can load the data in using any python SEGY library. You don't need to convert it to...

Judging from the error being raised, it looks like the horizon file isn't the size it should be. Is it possible that the downloaded horizon was corrupted or truncated in...