Using the loaded data#
native is the object that a dataset’s format library returns, and df is a pandas DataFrame. A dataset
has each only where it makes sense.
Each form is read the first time you use it, and kept, so using it again returns the same object.
Native objects#
from nbdemo.data.med import ecg_mitbih_119
record = ecg_mitbih_119.native # wfdb.Record
record.sig_name, record.fs
(['MLII', 'V1'], 360)
For a SigMF recording, native is a small SigMFRecording of our own: the metadata, and read_samples() for the
samples in the file beside it:
from nbdemo.data.rf import sigmf_synthetic
recording = sigmf_synthetic.native
iq = recording.read_samples()
iq.dtype, len(iq)
(dtype('complex64'), 1000000)
DataFrames#
ecg_mitbih_119.df.head()
| MLII | V1 | beat | |
|---|---|---|---|
| time | |||
| 0.000000 | -0.995 | -0.47 | NaN |
| 0.002778 | -0.995 | -0.47 | NaN |
| 0.005556 | -0.995 | -0.47 | NaN |
| 0.008333 | -0.995 | -0.47 | NaN |
| 0.011111 | -0.995 | -0.47 | NaN |
The ECG frame has a column for each lead, in physical units, indexed by seconds. The beat column holds
the annotation symbol at the sample of each beat. Flight telemetry is indexed by timestamp:
from nbdemo.data.aero import flight_dc
flight_dc.df.iloc[:3, :5]
| latitude | longitude | altitude | heading | track | |
|---|---|---|---|---|---|
| time | |||||
| 2026-01-01 12:00:00.000+00:00 | 38.8512 | -77.0402 | 5.0 | 10.0 | 0.0 |
| 2026-01-01 12:00:00.100+00:00 | 38.8512 | -77.0402 | 5.0 | 10.0 | 0.0 |
| 2026-01-01 12:00:00.200+00:00 | 38.8512 | -77.0402 | 5.0 | 10.0 | 0.0 |
When a form is not available#
A dataset with no such form raises FormUnavailable, which says why and lists the forms it has. The error
is an AttributeError, so hasattr also tells you:
print(hasattr(sigmf_synthetic, "df"))
print(sigmf_synthetic.forms)
try:
sigmf_synthetic.df
except AttributeError as error:
print(error)
False
('path', 'native')
rf/sigmf_synthetic has no df: the samples are binary IQ data, not a table; use .native, and its read_samples(). Its forms: path, native