search

dendrograms / astrodendro

Generate a dendrogram from a dataset
https://dendrograms.readthedocs.io/
Other
37 stars 38 forks source link

Viewer option for "integrated" datacube rather than datacube "slice" #95

Open tomr-stargazer opened 10 years ago

tomr-stargazer commented 10 years ago

When I'm working on the dendrogram of Tom Dame's CO survey, I often want to compare the structures identified by the dendrogram with the structures that were highlighted in these integrated maps: http://www.cfa.harvard.edu/mmw/MilkyWayinMolClouds.html

but the Viewer constrains me to look only at single slices (in velocity or latitude). Assuming that integrating along a single axis is a well-defined operation, would it be difficult to add an option that shows an "integrated" image rather than a single slice? The dendrogram computation should still use the full datacube, which is why I wouldn't want to feed Dendrogram an already-integrated array.

This may tie in neatly with Issue #94, wherein the best way to do this would be to write custom Viewers that take care of this.

astrofrog commented 10 years ago

Good idea - just to brainstorm, what would the contours show - the largest extent of the structure? I think for this case we'd have to disable clicking on the image to select the structure since there may be multiple structures along a line of sight.

tomr-stargazer commented 10 years ago

For a first pass, yes, I think "largest extent of the structure" would be the best way to do it, and then we can refine it from there if necessary.

I agree that clicking is ambiguous -- either disable it, or perhaps (if there's a clean way to do this) select all structures along that line of sight, and then narrow the selection further by clicking on the dendrogram or on a different view of the data? This requires supporting the selection of multiple structures of course.