features = {}
all_track_ids = list(df['song_id'].unique())
start = 0
num_tracks = 100
while start < len(all_track_ids):
tracks_batch = all_track_ids[start:start+num_tracks]
features_batch = sp.audio_features(tracks_batch)
features.update({ track_id : track_features
for track_id, track_features in zip(tracks_batch, features_batch) })
start += num_tracks
features['1mqlc0vEP9mU1kZgTi6LIQ']
tracks = pd.DataFrame.from_dict(features, orient='index')
tracks
I want to get the audio features of songs in spotify. But, I am trying to convert this dictionary into a dataframe, whose columns will be these features, for example valence, danceability and it shows this:
AttributeError Traceback (most recent call last)
in <module>
----> 1 tracks = pd.DataFrame.from_dict(features, orient='index')
2 tracks
in from_dict(cls, data, orient, dtype, columns)
1582 # TODO speed up Series case
1583 if isinstance(list(data.values())[0], (Series,
dict)):
-> 1584 data = _from_nested_dict(data)
1585 else:
1586 data, index = list(data.values()),
list(data.keys())
in _from_nested_dict(data)
10760 new_data: collections.defaultdict =
collections.defaultdict(dict)
10761 for index, s in data.items():
-> 10762 for col, v in s.items():
10763 new_data[col][index] = v
10764 return new_data
AttributeError: 'NoneType' object has no attribute 'items'
Now it shows:
|song_id|features |
|:------|:--------------------------------------:|
|link |{'valence:0.23', 'danceability:0.21',...|
How can it be like that?
|song_id|valence|danceability|
|link |0.23 |0.21 |
Can you help me please?
