For questions, use https://discourse.holoviz.org/ instead! This chat is for informal discussions and debugging sessions with developers of the HoloViz.org tools (HoloViews, Datashader, Panel, hvPlot, GeoViews, Param, etc.), but user questions should go to the HoloViz Discourse site so that others won't have to ask them again.
There's probably a transpose option to GridSpace as well, but I'm not sure about that.
Is it possible to plot a 2D slice in a 3D canvas with the plotly extension? Something like https://plotly.com/python/visualizing-mri-volume-slices/
Plotly 5.0 is coming. Don't know if changes anything for HoloViews or Panel. But something to note. https://community.plotly.com/t/plotly-py-5-0-is-coming-soon-try-the-release-candidate-today/53470
Hi, I am tring to develoip a little wms viewer I initially mixed bokeh and ipywidgets, but I am now convinced to move it to panel and pyviz - to start I am tring to follow this issue: holoviz/geoviews#481 to try to visualize WMS with
geoviews but I got stuck on the way I should load a standard WMS
I made a comment on the issue, attemting to explain my problem
Hi! I am trying to plot a multi-category boxwhisker plot with a scatter plot overlay. I am having some issues with getting this working. I am not sure how I should align the dimensions to get this working:
# example code
import math
import holoviews as hv
from holoviews import dim,opts
def changeOrderAndFont(plot,element):
factors = itertools.product(['F5','F4','F3','F2','F1'],['GroupA','GroupB'])
plot.state.x_range.factors = [*factors]
plot.handles['xaxis'].major_label_text_font_size = '12pt'
plot.handles['xaxis'].major_label_orientation = math.pi/2
plot.handles['xaxis'].group_text_font_size = '10pt'
plot.handles['yaxis'].axis_label_text_font_size = '10pt'
plot.handles['yaxis'].axis_label_text_font_style = 'normal'
plot.handles['xaxis'].axis_label_text_font_style = 'normal'
values = np.random.normal(0,1,size=50)
strata = np.random.choice(['GroupA','GroupB'],size=50)
feature = np.random.choice(['F1','F2','F3','F4','F5'],size=50)
df = pd.DataFrame({'Strata':strata,'Feature':feature,'Values':values})
pp = hv.BoxWhisker(df,kdims=['Feature','Strata'],vdims=['Values']).opts(cmap='Category20',box_color='Strata',height=400,width=400,hooks=[changeOrderAndFont])
# ss = hv.Scatter(df,kdims=['Feature','Strata'],vdims=['Values','Strata']) #gives a warning about a single kdim, and an error on dimension alignment
# ss = hv.Scatter(df,kdims=['Feature'],vdims=['Values','Strata']) #gives an error on dimension alignment
# (pp*ss) # errors here
ppAny help, suggestions, or direction would be appreciated.
oops just saw the header of the chat, will post this in the holoviews discourse site. sorry.
I would like to test and communicate the awesome Jupyter Lab button that is now available in pyviz_comms. But if I go to Panel binder https://mybinder.org/v2/gh/holoviz/panel/master?urlpath=lab/tree/examples I don't see it. And I have version 2.0.2 of pyviz_comms installed.
Is there anything I need to do?
You can presumably transform your data however you like, then add a custom axis with ticks and labels showing the untransformed values, but that would take some work. I don't think there's a simple way to do it at the Bokeh level without changes to Bokeh itself. If you want to have that flexibility added at the Bokeh level, you'd have to have a good motivating example for why log_e or configurability of log base more generally is compelling in a particular case. Most people who use log axes are interested in the overall axis-squashing behavior, which seems independent of base, not the precise mapping of specific values.
Yeah, I understand. That is what I had in mind as a fallback mechanism. It's indeed depends on domain science. In data science workflows one deals with counts and similar things - that's why log_10 works just fine. But I'm currently in chemistry domain now - and they speak log_e almost exclusively (reaction rates and other probability related entities). So they are interested in log_e scales. I would say that Chemistry is a big domain, especially given some recent bio-related issues in the world... I also would not know how easy it would be to add
configurable base to bokeh. And moreover, expose it through logy=True API of holoviews.
It would not be difficult to make it configurable in Bokeh and expose it in HoloViews. But you still do need to demonstrate why the log_e/log_10 difference matters, given that the axis is still labeled in un-transformed non-log units. How will a chemistry user even detect which one you used?
If you could manually make a plot and show the difference between the two logs, then that seems like a good enough demonstration to justify adding configurability over the log. Note that we recently (2020) generalized axis mapping in Bokeh to handle Datashader's nonlinear spacing, and it may already be feasible to support arbitrary transforms in a convenient way. Just make your demonstration on a Bokeh issue and ask Mateusz (a) how he'd do it right now, and (b) how hard it would be to make the log function be configurable as a simple option.