Dataset description
This dataset contains 1103 Python Numpy files (.npz), storing polarization profiles of 1103 FRBs which was used for Faraday rotation measure calculations in Xu et al., 2022, Nature, 609, 685.

1. Reading data and data content
The data files can be easily loaded using Numpy, for example:
>>> a = np.load('FRB20201124A_2021AprMay_FAST_Burst_0213.npz')
Each data object contains the following keys:
  I, Q, U, V: Stokes parameters of the profile
  SAT, BAT: Site arrival time and barycentric arrival time 
  pulse_range: Two integers indicating the left and right boundaries (bin indices) of the burst. 
  Tstart: UTC time of the first bin (MJD format)
  Tsamp: Sampling time per bin (seconds) 
  Frequency: Effective frequency of the profile (note: data are frequency-scrunched to a single channel; this value may differ slightly from FAST's nominal central frequency due to RFI removal)

The position angle of the linear polarization can be compuated as PA=arctan(U/Q)/2, also refer to Equation 1-2 in Jiang et al., 2022, RAA (https://doi.org/10.1088/1674-4527/ac98f6). 

Note on SNR and Re-binning, the Tsamp is kept same with the FAST raw data. For bursts with insufficient SNR, consider re-binning the data to improve per-bin SNR.

2. Burst indexing
Burst indices start at 1 and correspond to the indexing scheme in https://doi.org/10.6084/m9.figshare.19688854

Correction notice:
  Burst numbers in Fig. 2 of Xu et al. 2022 are corrected to: 795, 949, 1139, 1513.
  Burst numbers in Extended Fig. 4 are corrected to: 40, 383, 466, 571, 608, 741, 795, 1413, 1437.
