Skip to content

Carleton Astronet Admin Log Posts

IDL’s array index special handling, including the colon ( ‘:’ ) and the asterisk ( * )

Published by admin-joel on 2024-02-13

Note that IDL 2-dimensional arrays are given as Z[A,B] where A is column ref; B is row ref

x[*] means take each element of x (separately), so

x[*] = y[*] means,
x[0] = y[0], x[1] = y[1], … up to all elements of the arrays

x[0:2] is essentially does the same but on a subset of an array; eg
take elements 0 through 2 of the array (and do “whatever”)

I read also that this notation, esp on lhs, is much much faster than
conventional looping etc.

link example ln -s

Published by admin-joel on 2024-01-18

Here is a successful example of linking a file that really exists in subdir of PersArmNoIntarm,
up to a phake one in main PersArmNoIntarm.  Note I am doing it FROM desired location of
phake.  I say to ln -s the original, down in subdir, to “dot” which is current (main) dir:

PersArmNoIntArm> ln -s OutputFromLocBoots/PersNoIntArmvsLocSlopes.20240717_S.Summ.eps .

print idl’s path and see if it can find a file within it and note capitalization conventions

Published by admin-joel on 2024-01-02

IDL> print, file_which(‘chjuldaytolmstGALFA.pro’)
/usr/share/astro/idlshare/locallib/chjuldaytolmstGALFA.pro

 

and more specifically:

from https://stackoverflow.com/questions/29986130/idl-procedure-wont-compile

IDL procedure won’t compile

But when I try to find the procedure using “Findpro” I get

Procedure CHLOADCT found in directory  /data/clh93/colortables/CH/

So my paths are right. I don’t understand why it won’t find my procedure, does anyone know what’s going on?

Answers:

Things to check in situations like this:

  1. you are consistent about whether the routine is a function or procedure, i.e., routine is procedure and you are calling it as a procedure
  2. routine you are calling is either compiled already or the last routine in a file named the same as it with a .pro extension; the case of the filename and routine call must match or the filename must be all lowercase (just use all lowercase in filenames!)
  3. quick check to see if my_routine.pro is in your !path: 
    IDL> print, file_which('my_routine.pro')

    It will return an empty string if it can’t find it.

  4. make sure the file is not in your !path twice and you are getting someone else’s or an old version

Detect input (read) error without crashing program. Complicated fortran-like reads

Published by admin-joel on 2023-12-22

Fortran has the ability to send one somewhere upon an I/O error:

read(43,100,end=50) x,y,z         in which case program will go to statement labeled “50′ upon a read error.

IDL does not have such a simple way to do it. But after many hours, I got an equivalent process to work.  An example is in the
/data/psrdata/rmgalprojs/rm2016/AO_Pulsars_RM/Longi_RM_Plots.pro file, specifically in
the PersArmNoIntArm data input section.

——————————————————————————

The same section of code illustrates painful read in of a complicated data file, where a single (long) row had been turned into six. It is similar to but not quite identical to fortran. I had to eliminate all format = ‘(sd;jkfsd;jkg)’ statements because idl and i (≡fortran) did not interpret them the same. Instead I had to force variables to have the right type before reading data into them, by initializing them.  For example, to read the following variables properly, i did things like:

ztext = ‘  ‘
xfloat = 3.5
ydoubleprecision=8.d0

Then when I read a file into those variables from a file , idl will treat those variables as the types given in the initialization statements.

When Mirzam can’t be ssh-ed into (usually after reboot)

Published by admin-joel on 2023-10-16

When Mirzam can’t be ssh-ed into (usually after reboot):

/home and other NFS volumes exported by thuban2 probably aren’t mounted.
 
They should have automounted when mirzam was rebooted because they’re listed in /etc/fstab.

The fix is to login as root (which does not require /home) and then run the ‘mount -a’ command.

Bruce

Joel adds:  One needs root’s password in order to pull this off!

How to exit vncviewer’s fullscreen mode? full screen

Published by admin-joel on 2022-09-09

In Jan 2024:

click mouse inside window you want to take in or out of full screen – this gives this window primacy. E.g., vnc

hit <fn> f8 to bring up a menu with “full screen” option at top.*

Now let go of those two keys and that menu should persist

Now either click on full screen or else hit f key

This should toggle in and out of full screen mode.

*If menu is flashing, while keeping the <fn> f8 keys held down, also enter f (yes hand-gymnastics!)

====================================================================

On Sep 9, 2022 I had gave  the following two solutions. Don’t seem to work as of Jan 2024

To exit from full-screen mode, press Ctrl-Alt-Shift-F.

Alternately, press Ctrl-Esc Esc and then right-click on the vncviewer taskbar icon to see the menu.

where files really are on thuban2: /var/share (But not so fast for /usr/share/astro/ !)

Published by admin-joel on 2022-08-04

e.g.  /var/share/data/psrdata

–joel 20220804

(But not so fast for mirzam’s /usr/share/astro/ “real” location on thuban2:

when I did a “mount” command on mirzam on 20231231, I discovered that

/usr/share/astro/ on mirzam is really on thuban2’s /var/share/astro_centos7/    !  )

–joel 20231231

tl;dr:  this stems from our upgrade from centos6 to 7. The directory still occupying thuban’s
/var/share/astro/ is a carcass from centos6 days!