LoadFITS dialog.
Table of Contents
| Name | Direction | Type | Default | Description |
|---|---|---|---|---|
| Filename | Input | list of str lists | Mandatory | The name of the input file (note that you can give multiple file names separated by commas). Allowed extensions: [‘.fits’, ‘.fit’] |
| OutputWorkspace | Output | Workspace | Mandatory | |
| LoadAsRectImg | Input | boolean | False | If enabled (not by default), the output Workspace2D will have one histogram per row and one bin per pixel, such that a 2D color plot (color fill plot) will display an image. |
| FilterNoiseLevel | Input | number | 0 | Threshold to remove noisy pixels. Try 50 for example. |
| BinSize | Input | number | 1 | Rebunch n*n on both axes, generating pixels with sums of blocks of n by n original pixels. |
| Scale | Input | number | 80 | Pixels per cm. |
| HeaderMapFile | Input | string | A file mapping header key names to non-standard names [line separated values in the format KEY=VALUE, e.g. BitDepthName=BITPIX] - do not use this if you want to keep compatibility with standard FITS files. |
Load FITS files, which typically contain images, into a WorkspaceGroup. FITS stands for Flexible Image Transport System, see http://en.wikipedia.org/wiki/FITS. A new workspace (of type Workspace2D) is created for every FITS file loaded with this algorithm, and these workspaces are added into a WorkspaceGroup. The group workspace is named as indicated in the OutputWorkspace input property. The workspaces included in the group are named with the same name and an appendix _1, _2, etc., incremented sequentially as new files are loaded with the same OutputWorkspace property.
The way image pixels are loaded into the resulting workspaces depends on the porperty LoadAsRectImg. If it is set as false, one spectrum will be created for every image pixel. Otherwise, one spectrum will be created for every image row.
When LoadAsRectImg is true, the workspaces created by this algorithm contain one spectrum per row (and one bin per column). The first spectrum corresponds to the first (topmost) row and the last spectrum corresponds to the last (bottom) column. With this type of workspace you can display the image with a “color fill plot” or plot2D.
When LoadAsRectImg is false, the workspaces created by this algorithm contain one spectrum per pixel. The first spectrum corresponds to the top-left corner and the last spectrum to the bottom-right corner. The image pixels are assigned to spectra row by row, i.e., the first spectra correspond to the first or topmost row, the next spectra correspond to the next rows, going from top down, until finally the last spectra correspond to the last or bottom row.
The current implementation of this algorithm only supports 2D files (FITS images with two axes). With this condition, in principle this algorithm should be able to load any standard FITS file, and also those who do not deviate too much from the standard. See below an explanation of the minimum set of basic standard header entries that this algorithm requires (those should be present in any FITS file), So far, the algorithm has been tested with files produced by the cameras used for the IMAT instrument at ISIS, and Starlight Xpress CCD cameras used for calibration of other instruments at ISIS.
At a very minimum, the standard header entries SIMPLE, BITPIX, NAXIS, NAXIS1, and NAXIS2 must be present in the file.
This algorithm interprets extension headers defined for the IMAT instrument (ISIS facility). The set of extension headers for the IMAT instrument is being defined as of this writing and specific support and/or functionality related to additional headers might be added in this algorithm.
This algorithm uses one child algorithm:
ws_name = 'FITSimgs'
wsg = LoadFITS(Filename='FITS_small_01.fits', LoadAsRectImg=1, OutputWorkspace=ws_name)
ws = wsg.getItem(0)
# A couple of standard FITS header entries
bpp_log = 'BITPIX'
try:
log = ws.getRun().getLogData(bpp_log).value
print("Bits per pixel: {}".format(log))
except RuntimeError:
print("Could not find the keyword '{}' in this FITS file".format(bpp_log))
axis1_log = 'NAXIS1'
axis2_log = 'NAXIS2'
try:
log1 = ws.getRun().getLogData(axis1_log).value
log2 = ws.getRun().getLogData(axis2_log).value
print("FITS image size: {} x {} pixels".format(log1, log2))
print("Number of spectra in the output workspace: {}".format(ws.getNumberHistograms()))
except RuntimeError:
print("Could not find the keywords '{}' and '{}' in this FITS file".format(axis1_log, axis2_log))
Output:
Bits per pixel: 16
FITS image size: 512 x 512 pixels
Number of spectra in the output workspace: 512
ws_name = 'FITSws'
wsg = LoadFITS(Filename='FITS_small_01.fits', OutputWorkspace=ws_name)
ws = wsg.getItem(0)
# A couple of standard FITS header entries
bpp_log = 'BITPIX'
try:
log = ws.getRun().getLogData(bpp_log).value
print("Bits per pixel: {}".format(int(log)))
except RuntimeError:
print("Could not find the keyword '{}' in this FITS file".format(bpp_log))
axis1_log = 'NAXIS1'
axis2_log = 'NAXIS2'
try:
log1 = ws.getRun().getLogData(axis1_log).value
log2 = ws.getRun().getLogData(axis2_log).value
print("FITS image size: {} x {} pixels".format(log1, log2))
print("Number of spectra in the output workspace: {}".format(ws.getNumberHistograms()))
except RuntimeError:
print("Could not find the keywords '{}' and '{}' in this FITS file".format(axis1_log, axis2_log))
Output:
Bits per pixel: 16
FITS image size: 512 x 512 pixels
Number of spectra in the output workspace: 262144
Categories: Algorithms | DataHandling\Imaging
C++ source: LoadFITS.cpp (last modified: 2017-09-09)
C++ header: LoadFITS.h (last modified: 2016-06-07)