RFI mitigation

SEEK’s RFI mitigation follows the Offringa et al. SumThreshold algorithm. It’s implemented in pure Python and JIT-compiled for speed with the HOPE package.

It can easily be used without of the SEEK data processing pipeline:

import numpy as np
from seek.mitigation import sum_threshold

rfi_mask = sum_threshold.get_rfi_mask(tod=np.ma.array(data),
                                chi_1=20,
                                sm_kwargs=sum_threshold.get_sm_kwargs(40, 20, 15, 7.5),
                                di_kwargs=sum_threshold.get_di_kwrags(3, 7))

The TOD has to be a Numpy masked array when passed to the sum_threshold algorithm. The other parameters are optional an give you the possiblity to tune the mitigation. Crucial is the a good starting value of chi_1, best explored by trial and error. The keyword-arguments control the smoothing and dilation process. Further options can be found in the documentation of the module.

The resulting boolean mask looks something like that