sc-los
Introduction
Autogenerate lines-of-sight and their corresponding \(\mathit{I, Q, U}\) data and their derivatives.
Outputs
Sc-los generates the following general outputs:
- A region file containing the selected lines of sight.
- Plot showing locations of the 'valid' regions across the source
NumPy
pickle files for each LoS containing its Stokes data.
Each pickle file (with the file extension *.npz
) stores the data in a N-dimensional array, where N is the number of available channels. Each file contains the following keys:
Key | Description |
---|---|
I | I data for that LoS |
Q | Q data for that LoS |
U | U data for that LoS |
i_mfs_noise | Noise from the reference I MFS image (same for all channels) |
q_mfs_noise | Noise from the reference I MFS image (same for all channels) |
u_mfs_noise | Noise from the reference I MFS image (same for all channels) |
I_err | RMS noise extracted from a specific region in across the I spectrum |
Q_err | RMS noise extracted from a specific region in across the Q spectrum |
U_err | RMS noise extracted from a specific region in across the U spectrum |
lpol | Linear polarised intensity per-channel |
lpol_err | Linear polarised intensity error per-channel |
fpol | Fractional polarisation per-channel |
fpol_err | Fractional polarisation error per-channel |
pangle | Polarisation angle per-channel |
pangle_err | Polarisation angle error per-channel |
mask | Mask boolean values. They are true (masked) where data is not sensible |
freqs | Frequencies per-channel |
As an example with a typical run using showrunner, the output appears like:
├── scrap-outputs # Outputs from Los data generators
│ ├── los-data # LoS data outputs generated by sc-los
| | ├── reg_1.npz
| | ├── reg_1.npz
| | ├── ...
| | ├── reg_n.npz
| |── regions # Contains all the region files generated during a run
| | ├── regions-size-3-default.reg # All regions fitting within the specified mask region
| | ├── regions-size-3-default.png # An illustration of the locations in regions-size-3-default.reg
| | ├── regions-size-3-default-valid-candidates.reg # Selected regions during the first pass of sc-los
| | ├── regions-size-3-default-valid-candidates.png # An illustration of the locations in regions-size-3-default-valid-candidates.reg
| | ├── beacons.reg # A copy of regions-valid.reg for PolarVis
| | ├── noise-region.reg # Region selected for RMS noise extraction
| | ├── regions-valid.reg # Region file with positions of the 'valid' lines-of-sight
| | ├── regions-valid.png # Figure showing the regions in regions-valid.reg across the source
The final and most important region file is regions-valid.reg
(also called beacons.reg
).
Help Menu
The help menu is available using the command:
sc-los -h
which results in the output below. However, as the tool is under active development, the help menu may change in the future and should be checked before use using the help command after installation.
usage: sc-los [-h] [-nri NOISE_REF] [-todo] [--noverwrite] [-o] [-t] [-j] [--debug] [-ro] [-mrn] [-rf] [-rs] [-m] [-lo] [-idir IMAGE_DIR | -cubes CUBES CUBES CUBES] [-freqs] [-ref-image WCS_REF]
[-nrf NRFILE] [-mft] [-psnr] [-po] [--plot-grid] [-piqu] [-pfp] [-plp] [--ymax] [--ymin] [--xmax] [--xmin] [-p [...]] [-ps [...]]
________________________________________________________________________________
Description
Generate I, Q and U data for various LoS from image cubes.
This script uses the total intesity MFS image to determine regions
of the source with enough SNR and selects those as the lines of sight.
For now, this is tuned specifically for Pictor A, but it may be
extended and revised in the future.
The following could be generated from this script:
. 1. Region file containing the selected lines of sight
. 2. Plot showing locations of the regions on the source
. 3. Pickle files containing the following keys for the data:
. - I - Q - U
. - i_mfs_noise - q_mfs_noise - u_mfs_noise
. - I_err - Q_err - U_err
. - lpol - lpol_err
. - fpol - fpol_err
. - pangle - pangle_err
. - mask - freqs
. i.e for a single line of sight. Each LoS contains data for all
. the available channels
.
. 4. Optionaly, generate plots for fractional polarisation vs
lambda squared for each LoS
________________________________________________________________________________
options:
-h, --help show this help message and exit
-todo , --todo A string containing to do items. Specify using: (r): generate regions, (l): generate LOS data, (p): generate plots. Default is 'rl'
Options:
-nri NOISE_REF, --noise-ref-image NOISE_REF
The total intensity image used to get the global noise reference.
General arguments:
--noverwrite Do not ovewrite everything along the way. Default is overwrite
-o , -odir , --output-dir
where to dump output
-t , --threshold If SNR below which data will not be considered. Default is 3
-j , --nworkers How many workers to use for processing
--debug Disble parallel processing and enables sequential mode.
Region generation arguments:
-ro, --regions-only Only generate the region files. This requires the following options: --ref-image --mask --region-size
-mrn , --minimum-region-noise
Specific noise floor to generate the regions.
-rf , --region-file An input region file. Otherwise, one will be auto-generated. Genereated regions will be stored here
-rs , --region-size Create regions of this circle radius and perform analyses on them. If you want to set the data threshold, please use --threshold.
-m , --mask Mask containing the area where LoS should be restricted. This can be a FITS file or a region file (*.reg). It's REQUIRED for automatically making regions.
LoS Data generation arguments:
-lo, --los-only Only generate the line of sight data files. The following options should be specified: --idir/--cubes --ref-image --mask/--region-file --noise-ref-image --noise-
ref-file
-idir IMAGE_DIR, --image_dir IMAGE_DIR
Where the channelised I, Q and U images are
-cubes CUBES CUBES CUBES, --cubes CUBES CUBES CUBES
The I, Q, U image cubes (in this specific order) to be used. This will require specification of --freq-file
-freqs , --freq-file
Text file containing frequencies to be used. This is only active when the input FITS images are cubes and is particularly useful when the frequencies of images that form the
cube do not increase monotonically.
-ref-image WCS_REF, --ref-image WCS_REF
The reference image that will be used to generate the default region file. Must be the stokes I MFS image. This image will also be used to get the reference WCS for region
file generation.
-nrf NRFILE, --noise-region-file NRFILE
A region file containing region to be used as the noise reference.
-mft , --minimum-flag-threshold
Fraction of flags above which lines of sight should be ignored. Can be useful if you want to plot all the generated LOS. Otherwise, they will be filtered out. The simple
filter is that where values of fractional polarisation i are >1 or <0, this data is 'flagged'. behaviour: flag size > 0.7 of total data is flagged, ingore. Max is 1, min is
>0. Default 0.7
-psnr, --use-polzd-snr
Use to elect use of polarised SNR to determine valid LoS. Default S/N used is total intensity: rms noise
Plotting Arguments:
-po, --plots-only Only do plots
--plot-grid Enable to make gridded plots
-piqu, --plot-iqu Plot Q and U values
-pfp, --plot-frac-pol
Plot Fractional polarization
-plp, --plot-linear-pol
Plot linear polarization power
--ymax Y axis max limit
--ymin Y axis min limit
--xmax Y axis max limit
--xmin Y axis min limit
-p [ ...], --plot [ ...]
Make plots for these region sizes manually. These will be linearly scaled
-ps [ ...], --plot-scales [ ...]
Scales for the plots. Can be a space separated list of different scales. Options are linear or log.