Quickstart¶

This quick tutorial assumes that the nenupytf package is properly installed (see Installation).

Observation directory¶

NenuFAR/UnDySPuTeD high-rate time-frequency data files are usually gathered in one directory per observation. These files are of several different types although the only required ones are those labelled '*.spectra', each one corresponding to the one of the ouptut lanes of UnDySPuTeD (to a maximum of four although there may be less depending on the observation configuration).

For the rest of this tutorial, observation_XX is the observation name and the data are stored in '/path/to/observation':

$ ls /path/to/observation

observation_XX_0.spectra
observation_XX_1.spectra
observation_XX_2.spectra
observation_XX_3.spectra

Observation properties¶

Before running into the analysis, it may be usefull to understand what is inside these different lane files. The command-line nenupytf-info is here for that. It takes the path to the observation directory as argument:

$ nenupytf-info -o /path/to/observation

--------------- nenupytf ---------------
Info on /path/to/observation/observation_XX_0.spectra
Lane: 0
Time: 2019-11-28T19:01:45.0000000 -- 2019-11-29T02:59:34.7455616
Frequency: 11.9140625 -- 49.4140625 MHz
Beams: [0]
----------------------------------------

--------------- nenupytf ---------------
Info on /path/to/observation/observation_XX_1.spectra
Lane: 1
Time: 2019-11-28T19:01:45.0000000 -- 2019-11-29T02:59:34.7455616
Frequency: 49.4140625 -- 86.9140625 MHz
Beams: [0]
----------------------------------------

--------------- nenupytf ---------------
Info on /path/to/observation/observation_XX_2.spectra
Lane: 2
Time: 2019-11-28T19:01:45.0000000 -- 2019-11-29T02:59:34.7455616
Frequency: 11.9140625 -- 49.4140625 MHz
Beams: [1]
----------------------------------------

--------------- nenupytf ---------------
Info on /path/to/observation/observation_XX_3.spectra
Lane: 3
Time: 2019-11-28T19:01:45.0000000 -- 2019-11-29T02:59:34.7455616
Frequency: 49.4140625 -- 86.9140625 MHz
Beams: [1]
----------------------------------------

This command prints all the relevant informations about each lane file:

the path
the lane index
the minimal and maximal UT times
the minimal and maximal frequencies
the beam index/indices

This can also be retrieved within the Python interpreter (see Initialization).

Initialization¶

The Spectrum object is the most relevant to handle a given observation. In order to instanciate such object, it needs to be provided with a path to an observation directory (by default the current directory is used):

>>> from nenupytf.read import Spectrum
>>> spectrum = Spectrum('/path/to/observation')

Observation properties can also be printed directly:

>>> spectrum.info()

--------------- nenupytf ---------------
Info on /path/to/observation/observation_XX_0.spectra
Lane: 0
Time: 2019-11-28T19:01:45.0000000 -- 2019-11-29T02:59:34.7455616
Frequency: 11.9140625 -- 49.4140625 MHz
Beams: [0]
----------------------------------------

--------------- nenupytf ---------------
Info on /path/to/observation/observation_XX_1.spectra
Lane: 1
Time: 2019-11-28T19:01:45.0000000 -- 2019-11-29T02:59:34.7455616
Frequency: 49.4140625 -- 86.9140625 MHz
Beams: [0]
----------------------------------------

--------------- nenupytf ---------------
Info on /path/to/observation/observation_XX_2.spectra
Lane: 2
Time: 2019-11-28T19:01:45.0000000 -- 2019-11-29T02:59:34.7455616
Frequency: 11.9140625 -- 49.4140625 MHz
Beams: [1]
----------------------------------------

--------------- nenupytf ---------------
Info on /path/to/observation/observation_XX_3.spectra
Lane: 3
Time: 2019-11-28T19:01:45.0000000 -- 2019-11-29T02:59:34.7455616
Frequency: 49.4140625 -- 86.9140625 MHz
Beams: [1]
----------------------------------------

Data selection¶

Data selection can then be easily achieved with Spectrum.select(). The following example illustrates the selection of the first second of data from the beam 0 between 30 and 55 MHz (i.e. frequencies that are spead over the lanes 0 and 1. The result is stored in the spec variable.

>>> spec = spectrum.select(
        stokes='I'
        time=['2019-11-28T19:01:45.0000000', '2019-11-28T19:01:46.0000000'],
        freq=[30, 55],
        beam=0
    )

Averaging¶

Data can be averaged in time and frequency with Spectrum.average(). The following example shows how to average 5 min of data with a 0.1 MHz frequency resolution and a 0.1 sec time resolution:

>>> spec = spectrum.average(
        stokes='I'
        time=['2019-11-28T19:02:00.0000000', '2019-11-28T19:07:00.0000000'],
        freq=[30, 55],
        beam=0,
        dt=0.1,
        df=0.1
    )

Displaying the data¶

To quickly display the data:

>>> from nenupytf.display import plotdb
>>> plotdb(spec)