MontageError: Invalid table file

Question

I'm using Python 2.7.3 on my Debian server. Here is the input code:

import aplpy
import pyfits
from numpy import *
import matplotlib as plt
import montage as montage_wrapper
import PIL

r = pyfits.open('3c324IR.fits')
b = pyfits.open('3c324UVIS.fits')
g = pyfits.open('3c5GHZ.fits')

r.info()
g.info()
b.info()

print(r[1].header)
print(g[1].header)
print(b[1].header)

r1 = r[1].data
g1 = g[0].data
b1 = b[1].data

hdu = pyfits.PrimaryHDU(r1)
hdulist = pyfits.HDUList([hdu])
hdulist.writeto('r.fits')

hdu = pyfits.PrimaryHDU(g1)
hdulist = pyfits.HDUList([hdu])
hdulist.writeto('g.fits')

hdu = pyfits.PrimaryHDU(b1)
hdulist = pyfits.HDUList([hdu])
hdulist.writeto('b.fits')

aplpy.make_rgb_cube(['r.fits','g.fits','b.fits'], '3c_324_rgb.fits')

aplpy.make_rgb_image('3c_324_rgb.fits','3c_324_rgb.png')

f = aplpy.FITSFigure('3c_324_rgb.fits')

f.show_rgb()
f.save('3c_324_rgb2.png')

print 'END'

Everything works fine until I start the aplpy.make_rgb_cube() portion of the code. I know this because when I start it and cut that out it works through fine (until it realizes there's no output from this piece to continue). Here's the mess in the terminal:

Traceback (most recent call last):
  File "test9.py", line 47, in <module>
    aplpy.make_rgb_cube(['r.fits','g.fits','b.fits'], '3c_324_rgb.fits')
  File "/usr/local/lib/python2.7/dist-packages/aplpy/rgb.py", line 309, in make_rgb_cube
    montage.mMakeHdr(images_raw_tbl, header_hdr, north_aligned=north, system=system, equinox=equinox)
  File "/usr/local/lib/python2.7/dist-packages/montage_wrapper/commands.py", line 1468, in mMakeHdr
    return status.parse_struct("mMakeHdr", p.stdout.read().strip())
  File "/usr/local/lib/python2.7/dist-packages/montage_wrapper/status.py", line 33, in parse_struct
    result = Struct(command, string)
  File "/usr/local/lib/python2.7/dist-packages/montage_wrapper/status.py", line 70, in __init__
    raise MontageError("%s: %s" % (command, self.msg))
montage_wrapper.status.MontageError: mMakeHdr: Invalid table file: /tmp/tmpmyYyN7/images_raw.tbl

I don't know what the above means. I've looked through it for something to look around for in Google and I've come up empty. What is this error and how can I fix it?

Answer 1

aplpy.make_rgb_cube creates several files that are used by Montage mMakeHdr as a step along the way to generating your cube.
aplpy.make_rgb_cube gets past all the steps up to mMakeHdr.
mMakeHdr tries to read one of the created files named /tmp/tmpmyYyN7/images_raw.tbl Reading the file fails because the file ist invalid. An invalid file could be on with bad contents, could also mean that the file doesn't exist.

1. Did /tmp/tmpmyYyN7/images_raw.tbl actually get created?
2. If it is created, does it contain a list of image file names?
3. Check the creation date and see if this is the file you created or an older one that you can't overwrite.
4. Check that you have read and write permissions to /tmp and /tmp/tmpmyYyN7
5. If the file isn't there, but you have write permission to the temporary folder, check and see if a file named images_raw.tbl was written anywhere on your harddrive.

6. Check if you have more than one Montage installation on your system: Run "which mProjectPP" from a shell command line

   In Python do: import os os.system('which mProjectPP') Both should deliver the same answer. If not, you've got two Montage installations, and they are getting mixed up.

Answer 2

1. I went to /tmp/tmpmyYyN7/ and images_raw.tbl is there.
2. If I type "nano images_raw.tbl" it contains nothing...does that mean it doesn't contain any image file names?
3. I changed read/write permissions with "chmod 777" and it still didn't work when I ran it, just resulted in the same output.

4. typing "stat images_raw.tbl" results in the following:

   root@jarvis:/tmp/tmpmyYyN7# stat images_raw.tbl File: `images_raw.tbl' Size: 0 Blocks: 0 IO Block: 4096 regular empty file Device: 801h/2049d Inode: 5772965 Links: 1 Access: (0644/-rw-r--r--) Uid: ( 0/ root) Gid: ( 0/ root) Access: 2014-08-15 04:40:25.433245271 -0400 Modify: 2014-08-15 04:40:25.429245244 -0400 Change: 2014-08-15 04:40:25.429245244 -0400 Birth: -

5. This is the output:

   root@jarvis:/tmp/tmpmyYyN7# which mProjectPP /home/python_lib/Montage_v3.3/Montage//mProjectPP

Answer 3

Here is the output using the debugger:

> /home/python_lib/test10.py(59)<module>()
-> aplpy.make_rgb_cube(['/tmp/tmpmyYyN7/r.fits','/tmp/tmpmyYyN7/g.fits'], '3c_324_rgb.fits')
(Pdb) n
MontageError: MontageE...aw.tbl',)
> /home/python_lib/test10.py(59)<module>()
-> aplpy.make_rgb_cube(['/tmp/tmpmyYyN7/r.fits','/tmp/tmpmyYyN7/g.fits'], '3c_324_rgb.fits')
(Pdb) n
--Return--
> /home/python_lib/test10.py(59)<module>()->None
-> aplpy.make_rgb_cube(['/tmp/tmpmyYyN7/r.fits','/tmp/tmpmyYyN7/g.fits'], '3c_324_rgb.fits')

Then when I hit "n" "enter" again it shoots out the error:

(Pdb) n
Traceback (most recent call last):
File "test10.py", line 59, in <module>
aplpy.make_rgb_cube(['/tmp/tmpmyYyN7/r.fits','/tmp/tmpmyYyN7/g.fits'], '3c_324_rgb.fits')
File "/usr/local/lib/python2.7/dist-packages/aplpy/rgb.py", line 309, in make_rgb_cube
montage.mMakeHdr(images_raw_tbl, header_hdr, north_aligned=north, system=system, equinox=equinox)
File "/usr/local/lib/python2.7/dist-packages/montage_wrapper/commands.py", line 1468, in mMakeHdr
return status.parse_struct("mMakeHdr", p.stdout.read().strip())
File "/usr/local/lib/python2.7/dist-packages/montage_wrapper/status.py", line 33, in parse_struct
result = Struct(command, string)
File "/usr/local/lib/python2.7/dist-packages/montage_wrapper/status.py", line 70, in __init__
raise MontageError("%s: %s" % (command, self.msg))
montage_wrapper.status.MontageError: mMakeHdr: Invalid table file: /tmp/tmpQILi76/images_raw.tbl
Exception AttributeError: "'NoneType' object has no attribute 'path'" in <function _remove at 0x7f7ad03c2668> ignored

I'll look it over and see if I can make sense of it...I inserted the

pdb.set_trace()

line just before

aplpy.make_rgb_cube

and sure enough, that's where the error is. Is it possible that it's because of the way I've "sliced" the images in the previous code? The problem with the .fits images I'm running into is that there's multiple "image" files to each one:

Filename: 3c324IR.fits
No.    Name         Type      Cards   Dimensions   Format
0    PRIMARY     PrimaryHDU     748   ()
1    SCI         ImageHDU       144   (1091, 966)   float32
2    WHT         ImageHDU        48   (1091, 966)   float32
3    CTX         ImageHDU        43   (1091, 966)   int32
4    HDRTAB      BinTableHDU    557   2R x 274C    [9A, 3A, J, D, D, D, D, D, D, D, D, D, D, D, D, J, 6A, 9A, 7A, 18A, D, D, D, D, D, D, D, D, D, D, D, D, D, D, D, D, D, D, J, J, D, D, D, D, D, D, J, J, 8A, 23A, 11A, 18A, 4A, D, D, J, J, D, D, D, D, 23A, D, D, D, D, J, J, D, 23A, 8A, J, D, D, D, 23A, 1A, D, D, D, D, D, D, 12A, 12A, 8A, 23A, D, D, 10A, 10A, D, D, D, 2A, 3A, 3A, 4A, 8A, 7A, D, J, D, 6A, 9A, D, D, D, 4A, 18A, 3A, J, 5A, D, D, D, 8A, D, D, D, D, D, 23A, 1A, D, 23A, D, D, D, 3A, J, 1A, 4A, D, 3A, 6A, D, D, D, D, D, 23A, D, D, D, D, D, 1A, J, J, J, J, 8A, 23A, J, J, 10A, 7A, D, D, D, D, D, D, D, D, D, D, D, D, D, D, D, D, D, 12A, D, 24A, 23A, D, 2A, 8A, D, J, D, D, 6A, 4A, D, D, D, D, D, D, D, D, D, D, D, 19A, D, J, 23A, D, 8A, D, D, 13A, D, D, J, D, D, 5A, 9A, 1A, 4A, D, 1A, J, 1A, 1A, 1A, 1A, D, D, D, D, D, D, 4A, D, D, 4A, 3A, 1A, J, 3A, 6A, J, D, D, D, 4A, J, D, D, D, J, J, J, 1A, D, D, D, D, J, J, 6A, J, 8A, D, D, 1A, 5A, J, 3A, 8A, 8A, D, J, D, 8A, 8A]
5    WCSCORR     BinTableHDU     59   7R x 24C     [40A, I, 1A, 24A, 24A, 24A, 24A, D, D, D, D, D, D, D, D, 24A, 24A, D, D, D, D, J, 40A, 128A]

Above is the output when I print the Hubble IR .fits header, which shows there are several HDU's, but the only one I want is #1, which is the image of the field. The same follows for the UVIS and VLA images. The only way I could think to select one of them is by opening them with PyFITS, and then selecting each one with a line like

r1 = r[1].data

Is this the proper way to go about this? I guess most .fits images don't have this multi-image setup to one file. When I try running it the other way, open on the original file though it definitely doesn't like it.

After selecting with the line above I simply save it, individually, back as a .fits image.

Brandon Doyle