Closed lcolladotor closed 2 years ago
https://github.com/drighelli/SpatialExperiment/issues/25 is related since that's where the scalefactors_json.json
info would be stored at
@jvelezmagic @roramirezf maybe you have some ideas about this
I added options for controlling the point size (so the spot size) in the plots. Check it out at https://libd.shinyapps.io/spatialLIBD_Human_Lymph_Node_10x/ which so far is the only updated shiny app (requires spatialLIBD versions 1.6.4 or 1.7.6).
Here’s the default point size of 1.25
which we know looks quite bad when you zoom in
if you increase the point size to say 5
then without zooming in, it looks huge (almost like pop art)
but now it looks better when you zoom in
This issue isn't solved, but well, at least now we have a workaround available.
For the grids with Sang Ho @shkwon17, we found that point_size = 2
worked well for visualization when printing with the default height and width when using vis_grid_gene()
or vis_grid_clus()
. For example at https://github.com/LieberInstitute/Visium_IF_AD/blob/master/plots/initial_exploration/wholegenome_graph_based.pdf (private link right now).
The workaround has been working well, so I'll close this issue for now.
It'd be nice if we:
(a) made the point diameter use the actual spot diameter info from
SpaceRanger
See https://github.com/LieberInstitute/HumanPilot/blob/master/10X/151507/scalefactors_json.json for example where we have thefullres
diameter, and when taking into account the scaling factor, we should get the right saylowres
diameter.(b) update the
plotly
code such that the interactive plots allow you to Zoom in and rescale the spot diameters. This seems to be quite tricky based on threads like https://stackoverflow.com/questions/53217404/specifying-marker-size-in-data-unit-for-plotly, https://community.plotly.com/t/point-size-based-on-zoom/6788/2 and https://pretagteam.com/question/plotly-bubble-map-resize-markers-on-zoom.