View on GitHub

IRAF Community Distribution

IRAF maintained by the community

Home | Installation | Packages | X11IRAF | PyRAF | Forum

iraf imexamine crashes on macOS

Xin Wang wrote on Sep 15, 2019


Hi all,

I'm using pyraf that comes with astroconda ($ conda create -n iraf27 python=2.7 iraf-all pyraf-all stsci). The version of my system OS and software is as follows.

macOS Mojave 10.14.6

pyraf 2.2.dev

ds9 8.0.01

The problem is as follows.

$ pyraf

--\$this->_split2($m[0]) !ds9 some fits file

--\$this->_split2($m[0]) imexamine

Then I do see the mouse pointer in ds9 turned into a black circle. When I move the cursor on objects (inside a region circle), and press "r" to plot light profiles, the OS log itself out. When I input my login password, I see the following error message left on the pyraf terminal,

XIO: fatal IO error 35 (Resource temporarily unavailable) on X server "/private/tmp/com.apple.launchd.3HUq1QZZgt/org.macosforge.xquartz:0"

after 42063 requests (42063 known processed) with 0 events remaining.

Killing IRAF task `imexamine'

Traceback (innermost last):

File "", line 1, in

iraf.imexam(_save=1)

IrafError: Cannot open image display

I tried to switch from xgterm to xterm, but it did not help. Could any experts let me know how to fix this and make imexamine work nominally?

Thanks!

Xin

bernardowb wrote on Sep 24, 2019



Hi Xin!

The same is happening to me (as you, I'm using astroconda's pyraf on Mojave 10.14.6). The system log out also happens when I try to open an "epar" parameter windows...

Please, anyone could help us?

BWB

Mike Fitzpatrick wrote on Sep 24, 2019


The logout problem is a known bug in OSX (see https://github.com/matplotlib/matplotlib/issues/14999) for which I don't think there is yet a fix. This problem does not affect the classic CL but is triggered by the pyraf graphics window.

Xin Wang wrote on Sep 24, 2019


Quote by: fitzThe logout problem is a known bug in OSX (see https://github.com/matplotlib/matplotlib/issues/14999) for which I don't think there is yet a fix. This problem does not affect the classic CL but is triggered by the pyraf graphics window.


Hi Mike,

Thanks for your reply! Can I please ask you another question, also related to imexam? It works well on some fits images (pressing 'r', 'e', 's' etc. returns expected results), but on some other fits images, the pyraf terminal also returns the following.

Warning: Pixels out of bounds

and no plots pop up. (Sometimes, there's another piece of warning, "Warning: FXF: extname and/or extver value not found ('SCI')")

Do you know what's happening and how should solve this problem to get the radial profiles plotted?

Thanks!

Xin

Mike Fitzpatrick wrote on Sep 25, 2019


For OSX 10.14.6 you should be using DS9 v8.1b2 (see http://ds9.si.edu/site/Download.html).

In the past a lot of the "pixel subscript out of bounds" problems were related to invalid cursor reads that meant the computed image pixel was outside the size of the raster. This problem has shown up in various versions of DS9, is fixed, and then recurs again later. At the moment it is supposed to be fixed, and since you'll need the v8.1b2 version anyway chances are that installing the latest DS9 will also fix this problem.


Xin Wang wrote on Sep 26, 2019


Hi Mike,

I've updated ds9 to the v8.1b2 version but still experience the same problem. Do you have any further ideas about how to fix this?

Besides, I'm using an older OSX (10.11.6) to run imexam on pyraf, since imexam will crash on OSX 10.14.6. But I know through astroconda, one can specify some specific version of software to install (e.g., pip install photutils==0.6 installs version 0.6 instead of the current version 0.7). Do you know if this also works for pyraf and imexam? If so, which version should I choose to not make imexam crash on 10.14.6?

Thanks!

Xin

Quote by: fitzFor OSX 10.14.6 you should be using DS9 v8.1b2 (see http://ds9.si.edu/site/Download.html).

In the past a lot of the "pixel subscript out of bounds" problems were related to invalid cursor reads that meant the computed image pixel was outside the size of the raster. This problem has shown up in various versions of DS9, is fixed, and then recurs again later. At the moment it is supposed to be fixed, and since you'll need the v8.1b2 version anyway chances are that installing the latest DS9 will also fix this problem.



Last post on Sep 26, 2019