Python module to read SigMF datasets and categorize them into dataframes.
SigMF Directory: Path to the directory holding the SigMF dataset, can hold the dataset in multiple files
Pandas Dataframe: The output dataframe object for the entire dataset
Directory Loader: Enumerates all the SigMF recordings available at the provided path. Checks data consistency of each datafile to each meta datafile. Returns a dictionary of metafile and datafile.
Load metafile: Loads the elements of the metadata so the it can parse and enumerate the metadata contents.
Data Loader: Loads the datafile
Annotation Segmenter: Segments the datafile according to the annotations found in the metafile. Uses core::sample_start and core::sample_count. Further comments such as node id is added as an extra feature in the pandas data row for each annotation
Pandas Exporter: Combines the the segments for each annotated capture from different recording into a single dataframe
from sigmfreader import sigmfreader
# Define the number of samples in one timeseries training samples
ncols=128
# Call the reader
reader=sigmfreader("SigMF_dir", ncols)
# Do the pd exporter
df = reader.pandas_exporter()
# print the dataframe
df.head()Not generalized yet to handle all sigMF Json name/values, will keep updating to make it more generalized