clocktec

Name

lofar-clocktec

Description

Takes lofar calibration phases and do clock/tec separation

Installation

Use pip:

pip install lofar-clocktec

You can also directly install from the repository:

pip install git+https://git.astron.nl/RD/clocktec.git

Usage

Here is an example in how to use the code

import numpy as np
from lofar_clocktec.io import CTInput
from lofar_clocktec.fit_clock_tec import fit_clock_tec
from astropy import units as u
from astropy.time import Time

ct_input = CTInput(
    phases=phases[:, :, :] * u.rad,  # shape of (times, freq, stations) in rad
    freqs=freq * u.Hz,  # in Hz
    flags=flags[:, :, :],  # shape of (times, freq, stations)
    stations=stations[:],  # np array with station names
    times=Time(times / 86400, format="mjd", scale="utc"),  # convert MJD seconds to Time
)

result = fit_clock_tec(
    ct_input,
    ref_station=1,  # CS002 usually
    num_params_init=2,
    spatial_fit=True,  # to get improved reliable phase offsets
    lofar2_mode=True,  # if lofar 2 data - searches for smaller delays for RS
)

tec = result.tec  # TECU
delay = result.clock  # ns
offset = result.phase0  # rad (can be multiples of 2pi)

The function parameters are:

ct_input: CTInput - the data to be fitted, see above hoe to define it

weight_mode: str = "kappa" - how to calculate the weights using the phases. default: kappa from the von mises distribution

initialize_mode: str = "robust" - how to initialize the values for the first time slots, default: robust, it is slower but most reliable

init_with_previous: bool = True - use the previous time slot as initial value (except the first one) - makes the code much faster, default: True

num_params_init: int | None = None - how many params to fit? (2 - delay/tec, 3 - delay/tec/phase_offset, 4 - delay/tec/phase_offset/3rd tec). The most reliable phase offset is calculated with spatial_fit (see below), default: None (will be then be selected automatically - not always reliable yet)

final_ph0_fit: int = 0 - in progress, estimation of phase_offset with a different method - not reliable, default: False

remove_steps: bool = False - only for HBA, where the solutions can have jumps due to degeneracies. This fixes them, default: False

spatial_fit: bool = False - does two things: finds reliable phase offsets and fixes overall jumps for the HBA caused by degeneracies, default: False (recommended to set this to true if you need phase offsets)

allow_pi_wraps: bool = False - set True for IS where the Faraday correction is not good and the data can have pi jumps as well, default: False ref_station: int = 0 - the index of the reference station, default: 0

smooth_solutions: bool = False - smooths final solutions (useful for noisy data) - default:False

lofar2_mode: bool = True - searches for < 20ns delays for both CS and RS. Will fail on LOFAR1 data. Default: True.

log_file: By default a log file general_logs.log is written to the current working directory. Pass log_file=None to suppress this.

Authors and acknowledgment

Main authors: Cristina Cordun, Maaijke Mevius. Partially based on code by Michiel Brentjens

License

This project is licensed under the Apache License Version 2.0