bradenmacdonald / astrodendro Goto Github PK

For data cubes that use CGS units, densities may be on the order of 10^-21; consequently, the code as written in astrodendro/viewer/dendrowidget.py will see both the min and max data points as essentially 0, and the minimum_delta and minimum_flux parameters cannot be set to reasonable values.

In the interim, two workarounds are possible:

1. Once can manually set the parameters using the shell: for example, one can type make_dendrogram(10E-20, 20, 10E-21) or, more explicitly, make_dendrogram(min_flux=10E-20, min_npix=20, min_delta=10E-21) and it should spit out the desired dendrogram.

2. Scale the data. This can be done before using the viewer, or while the viewer is running, e.g.:

    import numpy as np
    set_data( np.log10(cube.data * 1E22) ) 
   

@bradenmacdonald - now that we have a release candidate for the merged astrodendro (formerly dendro-core) package, we should deprecate, remove, or rename our old repositories to avoid confusion.

The new URLs are http://dendrograms.org and https://github.com/dendrograms/astrodendro

I've renamed mine and changed the README.md:

https://github.com/astrofrog/astrodendro-deprecated

and will eventually remove it.

I got the following error after attempting to use dendrogram.to_hdf5:

  File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/Astronomical_Dendrograms-0.5.0-py2.6.egg/astrodendro/components.py", line 212, in newick
newick_items = [item.newick for item in self.items] 
RuntimeError: maximum recursion depth exceeded

I think I had a very large dendrogram. I attempted to change the maximum recursion level (default is 1000), but that resulted in a segmentation fault. I'm not sure what the best workaround is, but I think at least a try/except testing for a RuntimeError in the newick retriever might be useful.

Because IPython keeps changing their API, IPython version 0.11 exactly is required to get the viewer to run. In its absence, the viewer should still run, perhaps using PythonCommandWidget from astrocube/astrocubeview.py

This is more of an astrocube issue, but:

OBJECT and LINENAME FITS headers should not be required. Documentation should explain what these headers are not assume PPV cubes, as PPP cubes (from simulations etc.) can and are being used as well.