Using the paths#
path is the main file of a dataset, as a pathlib.Path. Use it to test a file loader against a real file.
from nbdemo.data.aero import flight_dc
flight_dc.path
PosixPath('/home/runner/work/nbdemo/nbdemo/src/nbdemo/data/aero/flight_dc/telemetry.csv')
Pass it to the loader you are testing. This one is pandas.read_csv:
import pandas as pd
pd.read_csv(flight_dc.path).head(3)
| time | latitude | longitude | altitude | heading | track | groundspeed | airspeed | vertical_speed | pitch | roll | throttle | engine_rpm | fuel | outside_air_temp | bus_voltage | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 2026-01-01T12:00:00.000000Z | 38.8512 | -77.0402 | 5.0 | 10.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0609 | -0.3120 | 12.3752 | 1042.1085 | 79.9999 | 21.9024 | 13.7609 |
| 1 | 2026-01-01T12:00:00.100000Z | 38.8512 | -77.0402 | 5.0 | 10.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0256 | -0.0949 | 11.9916 | 1015.2043 | 79.9997 | 22.0440 | 13.8233 |
| 2 | 2026-01-01T12:00:00.200000Z | 38.8512 | -77.0402 | 5.0 | 10.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0132 | 0.3382 | 12.2338 | 1015.1106 | 79.9996 | 22.0184 | 13.7712 |
Datasets with several files#
Some formats keep their data in more than one file. For ecg_mitbih_119, path is the header, and the
signals and annotations are beside it. paths lists all of them by name:
from nbdemo.data.med import ecg_mitbih_119
ecg_mitbih_119.paths
{'119.atr': PosixPath('/home/runner/work/nbdemo/nbdemo/src/nbdemo/data/med/ecg_mitbih_119/119.atr'),
'119.dat': PosixPath('/home/runner/work/nbdemo/nbdemo/src/nbdemo/data/med/ecg_mitbih_119/119.dat'),
'119.hea': PosixPath('/home/runner/work/nbdemo/nbdemo/src/nbdemo/data/med/ecg_mitbih_119/119.hea')}
The files are in one folder, so a loader that finds the other files from the main file works unchanged:
import wfdb
wfdb.rdrecord(str(ecg_mitbih_119.path.with_suffix(""))).sig_name
['MLII', 'V1']
Large files#
Using path or paths downloads whatever is missing. The .sigmf-data file of sigmf_synthetic is 8 MB,
so it downloads the first time, to the cache. is_local says whether a dataset’s files are all present:
from nbdemo.data.rf import sigmf_synthetic
print(sigmf_synthetic.is_local) # False until a path is used
meta = sigmf_synthetic.path # downloads the recording
print(sigmf_synthetic.is_local)
sorted(p.name for p in meta.parent.iterdir())
False
True
['synthetic.sigmf-data', 'synthetic.sigmf-meta']