merge_packed_time_series

lsst_efd_client.merge_packed_time_series(packed_dataframe, base_field, stride=1, ref_timestamp_col='cRIO_timestamp', fmt='unix_tai', scale='tai')

Select fields that are time samples and unpack them into a dataframe.

Parameters:
packed_dataframe : pandas.DataFrame

packed data frame containing the desired data

base_field : str

Base field name that will be expanded to query all vector entries.

stride : int, optional

Only use every stride value when unpacking. Must be a factor of the number of packed values. (1 by default)

ref_timestamp_col : str, optional

Name of the field name to use to assign timestamps to unpacked vector fields (default is ‘cRIO_timestamp’).

fmt : str, optional

Format to give to the astropy.time.Time constructor. Defaults to ‘unix_tai’ since most internal timestamp columns are in TAI.

scale : str, optional

Time scale to give to the astropy.time.Time constructor. Defaults to ‘tai’.

Returns:
result : pandas.DataFrame

A pandas.DataFrame containing the results of the query.