How to plot full screen img or without scroll bar

My previous workflow is to write code with python and analyze/preview data in Matlab. Recently I want to go jupyter notebook with all of them so I explore holoviews in pyviz. It has fewer features compared with Matlab so I hope someone can let me know if it can be realized in holoviews.

1. popup/fullscreen plot with interaction zoom.

   • There is no fullscreen/popup image to see the detail of data [EDIT:This is a good feature]
   • also, %matplotlib qt doesn’t work
   • I use hv.extension('bokeh', 'matplotlib', width=100) so the width is bigger, but for big image, the height is limited within scroll bar. It’s terrible to observe the detail [UPDATE: just click left side can expand image] [UPDATE: hv.extension('bokeh', 'matplotlib', width=100) + %%output size=200]

2. horizontal/vertical zoom with area selection by mouse.

   • I know how to zoom h or v (by wheel zoom) but mouse zoom with selection interval is convenient and precise. [UPDATE: put the mouse in the place of any axis you want to fix and expand around, it will only zoom that area, so it has the same effect with H/V zoom of Matlab]
   • ctrl + z to undo last zoom [EDIT:This is a good feature]

3. annoying legend

   • how to remove it? [UPDATE: %%opts Curve [ show_legend=False]]

Hope holoviews can be better choice than Matlab, for analyzing huge data! Thanks.

[jbednar]

1. popup/fullscreen plot with interaction zoom. There is no fullscreen/popup image to see the detail of data

Yes, it would be great to have a “fullscreen” button on Bokeh plots that would expand them into full-screen mode like a video player does. This would need to be implemented by Bokeh, not HoloViews, and there is already a Bokeh feature request for that. That issue was opened in 2016, so it’s obviously not a priority for the existing developers, but if you want to help it along you can try implementing it; that issue already links to some sample code.

also, %matplotlib qt doesn’t work

Right; I don’t think anyone has ever worked on Qt support for the HoloViews matplotlib backend, and I don’t expect anyone to do so in the foreseeable future.

I use hv.extension(‘bokeh’, ‘matplotlib’, width=100) so the width is bigger, but for big image, the height is limited within scroll bar. It’s terrible to observe the detail

I’m not sure what you are hoping to happen here, but if it’s just that you don’t want the inner scrollbar, you can click on the area to the left of the plot, along the left margin, to disable the scrollbar and put the entire output onto the page. Of course, you’ll still need to use the whole page’s scrollbar at that point if it’s a very tall plot; not much that can be done about that!

1. horizontal/vertical zoom with area selection by mouse. I know how to zoom h or v (by wheel zoom) but mouse zoom with selection interval is convenient and precise.

You can choose between Wheel Zoom or Box Zoom, as you prefer:

Box zoom sounds like what you want here.

1. annoying legend how to remove it?

You can use .options(show_legend=False) to turn off the legend, but then you’ll have trouble figuring out which curve is which. Or you can just move it to a different location, using legend_position.

[jbednar]

@kitaev-chen, there’s a way to set which tool should be active by default, but it’s a bit clunky; see #2730.