These are chat archives for ipython/ipython

19th
Oct 2015
bsdis
@bsdis
Oct 19 2015 21:04
hi, i just want to hear your opinion on this. If i have eg, a large matrix i normally visualize using matplotlib's imshow, I can zoom and pan around in the image very smoothly. Will I be able to this with ipython notebook as well?
Dave Hirschfeld
@dhirschfeld
Oct 19 2015 21:09
@bsdis - if you run %matplotlib notebook in one of your cells it will set up the matplotlib webagg backend which allows for such interactivity in a notebook/web context.
bsdis
@bsdis
Oct 19 2015 21:12
@dhirschfeld aha i see.....how would that work for large matrices or plots with a lot of datapoints? I mean how does it scale on amounts of data compared to if it is run on a desktop?
Dave Hirschfeld
@dhirschfeld
Oct 19 2015 21:18
I'm not sure what you mean by how does it scale if you're concerned about the performance of the webagg backend that will probably depend on your system and the size of the data you're plotting. The only way to answer that would be to give it a go and see if it's acceptable for your particular usecases.
Dave Hirschfeld
@dhirschfeld
Oct 19 2015 21:24
IIUC it works by sending updated images which may not be particularly performant in which case you may want to check out bokeh which I think is aiming to be a performant interactive plotting solution for large data
bsdis
@bsdis
Oct 19 2015 21:26
i see :) ok thank you so much