Derivative features: The tempogram
One benefit of cleaning up your data is that it lets you compute more sophisticated
features. For example, the envelope calculation you performed is a common technique
in computing tempo and rhythm features. In this exercise, you'll use librosa
to compute some tempo and rhythm features for heartbeat data, and fit a model once more.
Note that librosa functions tend to only operate on numpy arrays instead of DataFrames,
so we'll access our Pandas data as a Numpy array with the .values attribute.
Cet exercice fait partie du cours
Machine Learning for Time Series Data in Python
Exercice interactif pratique
Essayez cet exercice en complétant cet exemple de code.
# Calculate the tempo of the sounds
tempos = []
for col, i_audio in audio.items():
tempos.append(lr.beat.____(i_audio.values, sr=sfreq, hop_length=2**6, aggregate=None))
# Convert the list to an array so you can manipulate it more easily
tempos = np.array(tempos)
# Calculate statistics of each tempo
tempos_mean = tempos.____(axis=-1)
tempos_std = tempos.____(axis=-1)
tempos_max = tempos.____(axis=-1)