The routine eis_spike_viewer is a widget-based routine for browsing data-sets that have been cleaned of cosmic rays. 

To use it do the following:

IDL> data=obj_new('eis_data',filename)
IDL> wd=data->getvar(iwin)
IDL> eis_debias,wd
IDL> eis_despike,wd,wd_clean,swtch=spikes
IDL> eis_spike_viewer,wd,wd_clean,spikes

Note that I'm using the routine eis_debias to remove the CCD bias. This is a routine I wrote (and is attached to this message). I think it does the same thing as the eis_prep version (removes the median of the lowest 2% pixels) but I don't know how to easily call this from the command line.

You should see a widget appear with 6 windows.

The top left window shows the original image, the top middle shows the cleaned image, and top right shows the positions of CRs (should just be a few dots).

You can choose 'Exposure' or 'Image'. The latter is for showing exposures (lambda-Y), while the former shows X-Y images.

The 'Zoom', 'Pixel' buttons:  when the former is set, you can zoom into any of the three images. Use the left mouse and click-and-drag on the image to produce a rubberband box. When you let go of the mouse button you will zoom into your selected region.  When 'Pixel' is set you can click on any pixel in any of the images and three new plots will appear on the bottom three graphics windows. These are...

Bottom left: a slice through the data-cube in the wavelength direction, with your selected pixel highlighted.

Bottom middle:  a slice in the exposure direction (X).

Bottom right: a slice in the Y direction.

In each window, the pixels that have been removed should be shaded white.

The slider allows you to browse through the data-cube: if lambda-Y images are shown, then it will flick through X values; similarly if X-Y images are shown then it will flick through lambda values.

The Step button's quite neat as it automatically steps through each exposure/image. The [] button was supposed to stop this process, but dosen't seem to work.

I had a quick look at one data-set and I didn’t spot any major problems with real data being removed, but it would be good if other people had a play with it.  If you do find something strange then please e-mail me with the details of the data-set.


Peter Young