Output
This section describes the output produced by nemo, and the main files of interest.
Directory Structure
The main nemo command creates a folder in the current working directory
under which it stores all of its output. By default this directory has the same name
as the configuration file, but with the .yml
extension removed. We will refer to
this as the output
directory hereafter.
After nemo has run, within the output
directory, you will find the
following:
output/
diagnostics/
filteredMaps/
selFn/
output_optimalCatalog.csv
output_optimalCatalog.fits
output_optimalCatalog.reg
- diagnostics/
This directory contains output that is useful for debugging (mainly plots in
.pdf
and.png
format, but the filter kernels are also cached here).- filteredMaps/
This directory contains the filtered maps, named according to the filter definitions given in the configuration file. If you have chosen to break up a large map into tiles for parallel processing (see Tiling), then this directory will contain subdirectories named after each tile. If
stitchTiles: True
is set in the config file (see Output Options), then you will find monolithic maps namedstitched_*.fits
in this directory. For more on the maps themselves, see Filtered Maps below.- selFn/
This directory contains all the necessary information for calculating the completeness of a cluster search, given a model for the relationship between the SZ signal and mass, and an assumed halo mass function. For both point source and cluster searches, you will find the survey area masks and noise maps in this directory.
- output_optimalCatalog.fits, output_optimalCatalog.csv
These files contain the object catalogs. Both contain the same information, but in different formats. The CSV file is plain text, but despite the extension is tab-delimited. Largely this is provided for ‘historical’ reasons, as the FITS table format file is more convenient (it can be read by both TopCat and Astropy Table without any special handling). For information on the contents of these catalogs, see Catalogs.
- output_optimalCatalog.reg
This is a region file that can be loaded into SAOImage DS9 to overplot the object locations on any FITS image.
Catalogs
The catalogs produced by nemo may contain the following columns:
- name
The name of the object.
- RADeg
The right ascension of the object in decimal degrees (J2000 typically, but this depends upon the WCS of the maps used with nemo).
- decDeg:
The declination of the object in decimal degrees (J2000 typically, but this depends upon the WCS of the maps used with nemo).
- SNR:
Signal-to-noise ratio, optimized over all filter scales.
- fixed_SNR:
Signal-to-noise ratio at the reference filter scale, chosen using
photFilterLabel
in the config file (see Cluster Mass Estimates).- numSigPix:
The number of pixels above
thresholdSigma
as set in the config file (see Object Detection and Photometry).- template:
Name of the matched filter template resulting in the highest SNR detection of this object.
- tileName:
The name of the tile in which the object was detected.
- galacticLatDeg:
Galactic latitude of the object in decimal degrees (can be used for pruning objects close to the Galactic plane, where contamination is naturally expected to be higher).
- deltaT_c:
The amplitude of the source, in ΔTemperature (μK) with respect to the CMB.
- err_deltaT_c:
The 1-sigma uncertainty on the amplitude of the source, in ΔTemperature (μK) with respect to the CMB.
- fluxJy:
The flux density of the source in Jy. Note this quantity is derived from
deltaT_c
, and depends upon the beam solid angle and the assumed effective frequency for the map.- err_fluxJy:
The 1-sigma uncertainty on the flux density of the source, in Jy.
- y_c:
Central Comptonization parameter (10-4) measured using the optimal matched filter template (i.e., the one that maximizes SNR).
- err_y_c:
The 1-sigma uncertainty on the central Comptonization parameter (10-4) measured using the optimal matched filter template (i.e., the one that maximizes SNR).
- fixed_y_c:
Central Comptonization parameter (10-4) measured at the reference filter scale, chosen using
photFilterLabel
in the config file (see Cluster Mass Estimates). This quantity is called ỹ0 in the ACT cluster papers.- err_fixed_y_c:
The 1-sigma uncertainty on the central Comptonization parameter (10-4) measured at the reference filter scale, chosen using
photFilterLabel
in the config file (see Cluster Mass Estimates).- tileBoundarySplit:
If
True
, this object may have been de-blended across a tile boundary. At present, the only way to determine if this really is the case is by visual inspection of the maps and catalogs (it happened rarely when using the ACT DR5 maps).
Filtered Maps
The nemo command produces two different types of filtered maps, assuming that you have configured nemo to write filtered maps to disk (see Filters):
- *_filteredMap.fits:
These are maps in signal units, set by
outputUnits
in the config file (see Filters). For point source searches, these will be ΔTemperature (μK) with respect to the CMB (ifoutputUnits: uK
is set in the config file). For cluster searches, these will be in terms of the central Comptonization parameter, y0 (ifoutputUnits: yc
is set in the config file).- *_SNMap.fits:
These are maps in signal-to-noise (S/N) units.