-
Notifications
You must be signed in to change notification settings - Fork 10
Troubleshooting
Problems you can encounter when running the script. Is your error not in the list? Even with the error in the list you can't figure it out? Send an email or make start an 'Issue' on the binoculars github page.
The computer cannot handle the size of the space object. Either there is something wrong in calculating the projections, or the resolution is to big. Evaluate with rerunning the script at a larger bin size.
For some reason the value of the monitor measured in either of the two hutches indicates zero. In all likelyhood you forgot to open the shutter. This means there will be not data in the images. Remove the scan from your selection. You can check the images manually by opening the scanviewer.
The filename you specfied does not exist, check if the filename is correct.
Something went wrong in the backend. Either there is a wrong specfication in the configuration file, the input data contains bugs or there is a bug in the code. They frequency of problems is usually in that order. Try to use the information in the error to pinpoint the error.
A specification in the configuration file went awry. Either the script can't parse the option you set in the configuration file, or that option clashes with the variouse other scripts. Like an an invalid range for the masks. First check if the setting can be parses properly. Second, check if the value makes sense.
The file explorer on nice can be slow and missing the new files. Check in the terminal of the files are not there, or byt opeing the in the gui.
Common problems when with features in the mesh.
Assuming you made the mesh using zap lines, the rings are due to background noise. If there is a scattering source not originating from the sample like a window or air, this results in a background which is dependant on the position of the detector. In other words, pixels that have the same cooridinate in reciprocal space can have a different value for the background. The specific pattern you will see depends on the amount of overlap of reciprocal space you have.
Check the value for the central pixel, did you swap x,y? Otherwise the wrong UB matrix might be specified in the specfile. Specify the UB matrix in the input section in the config file 'UB = 1, 2, 3, 4, 5, 6, 7, 8, 9'
Either there is no datapoint, or the value of the datapoint is zero. When using a maxipix the area between the chips is usually masked. This could in a white strip in your mesh, depending on the settings. When the value of the datapoint is zeros and is plotted with a logarithmic colorscale this datapoint is discarded.
- If you see your mesh needs background subtraction, make sure you measure the background. This makes it much easier to perform the background subtraction than to come up with a very complicated data treatment algorithm.
- Try to avoid modifying the structure of your images folder. This destroys the reference between the specfile and the locations of the images. This means you will have to manually enter the location to the images for every separate scan.
- Compress your images with gzip. Gzipped images can be read with binoculars and pymca. This means that you can leave the images compressed and there is no need to carry around a compressed and an uncompressed version of your images.