pinta

pinta is a data analysis pipeline for upgraded GMRT pulsar data. It RFI mitigates and folds uGMRT data to form PSRFITS archives.

A detailed description of pinta is given in Susobhanan et al., 2020 (https://doi.org/10.1017/pasa.2021.12). If you use this pipeline in your research, please cite this paper, along with those for the two RFI excision tools that pinta heavily make use of (RFIClean: https://doi.org/10.1051/0004-6361/202040164, gptool:to be updated).

pinta is now available at GMRT as an observatory pipeline. See this user guide for instructions on how to access pinta at GMRT.

Installing

Installing Dependencies

Install the following pulsar packages by following the installation instructions given in their respective websites/repositories.

• tempo2 (https://bitbucket.org/psrsoft/tempo2/src/master/)
• dspsr (http://dspsr.sourceforge.net/)
• psrchive (http://psrchive.sourceforge.net/)
• RFIClean (https://github.com/ymaan4/RFIClean)
• gptool (https://github.com/chowdhuryaditya/gptool)
• ugmrt2fil (https://github.com/inpta/ugmrt2fil)

Install the python dependencies.

$ pip3 install parse astropy pyyaml --user

Permissions

One persistent issue that arises while analyzing data using a pipeline is maintaining correct permissions for the data files. The way we deal with this is to use one user group for all analysis using pinta. For example, we will use the pulsar group name in this guide. The following command should be run at the start of every session before running pinta. (Make sure that this group is created and you are added to it.)

$ newgrp pulsar

Installing pinta

Say you want to install pinta in /pinta/install/path/.

$ git clone https://github.com/inpta/pinta.git
$ cp pinta/pinta pinta/pinta*.py pinta/pinta.yaml /pinta/install/path/
$ chgrp pulsar /pinta/inst/path/*
$ chmod ug+r /pinta/inst/path/*
$ chmod ug+x /pinta/inst/path/pinta

Now add this line to your .bashrc file.

$ export PATH=$PATH:/pinta/install/path/pinta

Configuring pinta

The pinta configuration is stored in pinta.yaml. This file should be saved in the same directory as the pinta executable script. pinta.yaml looks like this:

    pinta:
        pardir: /pinta/par/dir/
        gptdir: /pinta/gpt/dir/
        rficconf: /pinta/rfic/dir/inpta_rficlean.flags
        group: pulsar

• pardir is the directory where the pulsar ephemeris (par) files (named like J1234+5678.par) are stored. These files are needed for folding.
• gptdir is the directory where the gptool configuration files (gptool.in.*) are stored. See examples directory for examples.
• rficconf contains the flags to be passed to RFIClean. See examples directory for example.
• group is the user group which has permissions to run pinta.

Running pinta

Input files

For each observation uGMRT produces two files - a binary raw data file and a timestamp file. Both are required for analysis.

Usage

pinta is invoked from the command line with the following syntax.

$ pinta [--help] [--test] [--no-gptool] [--no-rficlean] [--nodel] [--retain-aux] [--log-to-file] [--gptdir <...>] [--pardir <...>] [--rficconf <...>] <input_dir> <working_dir>

• <input_dir> is the directory where the input (raw data and timestamp) files are saved.

• <working_dir> is the directory where the output files will be created. For pinta to run, this directory should contain a file called pipeline.in containing the metadata required to do the analysis.

• --help prints this syntax on the screen.

• --test performs the input validation steps and prints out the commands without executing them. Used for debugging.

• --no-gptool runs pinta without gptool.

• --no-rficlean runs pinta without RFIClean.

• --nodel keeps the intermediate data products.

• --retain-aux keeps the auxiliary putput files produced by component programs.

• --log-to-file writes the stdout to a log file.

• --gptdir specifies a custom directory for gptool.in files. This is recommended.

• --pardir specifies a custom directory for par files.

• --rficconf specifies a custom RFIClean configuration file.

pipeline.in file

The uGMRT raw data file does not contain any metadata required to analyze it. So we provide it through an ASCII file named pipeline.in. Each row in pipeline.in represents a raw data file. The columns are as follows:

1. JNAME - The J2000 of the pulsar (eg. J2124-3358)
2. RawData - The raw data file name (not the full path)
3. Timestamp - The timestamp file name (not the full path)
4. Freq - The local oscillator frequency (MHz)
5. Nbin - The number of phase bins for folding (Set to -1 to use the default. The default is the largest power of 2 less than psr_period/Tsmpl.)
6. Nchan - Number of frequency channels
7. BandWidth - Observation bandwidth (MHz)
8. TSmpl - Sampling time of observation (s)
9. SB - Sideband (USB - Upper sideband / LSB - Lower sideband)
10. Npol - Number of polarization products (1 - Total intensity / 4 - Stokes IQUV). Only total intensity is implemented at present.
11. TSubint - Subintegration duration for folding (s)
12. Cohded - Whether the data has been coherently dedispersed (1 - Yes / 0 - No)

An example is given in the examples directory.