In [1]:
from lstm.midiHandlers import midi_archive as ma
from lstm.fileHandlers import composers as cmprs
from lstm.fileHandlers import dataset as ds
import pandas as pd
In [2]:
ma.build_all_meta(r"C:\Users\livb1\Desktop\Yolane\2020Spring\Capstone\music-classification-deep-generation\lstm\midi")

C:\Users\livb1\Desktop\Yolane\2020Spring\Capstone\music-classification-deep-generation\lstm\midi
Found 17844 files from 12 composers!
Loading midi files with 3 threads...
-------

Saving meta csv...
Meta CSV file saved!
In [3]:
metadf = pd.read_csv(".\lstm\midi\meta.csv")
In [4]:
display(metadf)

filename composer type tracks 4per_beat first_key_sig predicted_key_sig first_time_n first_time_d first_time_32nd ... D D# E F F# G G# A A# B
0 C:\Users\livb1\Desktop\Yolane\2020Spring\Capst... Bach 1 6 240 C D# 4.0 4.0 8.0 ... 57.0 90.0 14.0 96.0 0.0 75.0 89.0 10.0 105.0 0.0
1 C:\Users\livb1\Desktop\Yolane\2020Spring\Capst... Beethoven 1 14 240 C C 2.0 4.0 8.0 ... 11.0 40.0 54.0 35.0 53.0 105.0 8.0 15.0 0.0 3.0
2 C:\Users\livb1\Desktop\Yolane\2020Spring\Capst... Haydn 1 12 192 NaN C 3.0 4.0 8.0 ... 12.0 4.0 107.0 8.0 5.0 19.0 30.0 23.0 0.0 34.0
3 C:\Users\livb1\Desktop\Yolane\2020Spring\Capst... Bach 1 6 240 C D# 4.0 4.0 8.0 ... 14.0 40.0 0.0 10.0 0.0 12.0 16.0 0.0 24.0 0.0
4 C:\Users\livb1\Desktop\Yolane\2020Spring\Capst... Haydn 1 12 192 NaN F 3.0 4.0 8.0 ... 33.0 8.0 31.0 29.0 9.0 105.0 1.0 31.0 42.0 12.0
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
10958 C:\Users\livb1\Desktop\Yolane\2020Spring\Capst... Webster 1 8 96 D D 3.0 4.0 8.0 ... 142.0 0.0 13.0 0.0 243.0 19.0 0.0 89.0 3.0 3.0
10959 C:\Users\livb1\Desktop\Yolane\2020Spring\Capst... Webster 1 8 96 D D 3.0 4.0 8.0 ... 114.0 12.0 17.0 0.0 188.0 29.0 0.0 76.0 4.0 11.0
10960 C:\Users\livb1\Desktop\Yolane\2020Spring\Capst... Webster 1 8 96 D G 3.0 4.0 8.0 ... 109.0 30.0 14.0 0.0 162.0 24.0 2.0 83.0 1.0 7.0
10961 C:\Users\livb1\Desktop\Yolane\2020Spring\Capst... Webster 1 8 96 D D 3.0 4.0 8.0 ... 140.0 0.0 13.0 0.0 243.0 20.0 1.0 90.0 3.0 4.0
10962 C:\Users\livb1\Desktop\Yolane\2020Spring\Capst... Webster 1 8 96 D D 3.0 4.0 8.0 ... 124.0 0.0 15.0 0.0 213.0 25.0 0.0 79.0 4.0 4.0

10963 rows × 27 columns

In [5]:
composers = cmprs.get_composer_works(metadf)
In [6]:
display(composers)

works
composer
Bach 2869
Beethoven 1713
Chopin 1229
Handel 870
Haydn 769
Hays 1022
Thomas 1116
Webster 1375
In [7]:
dataset = ds.VectorGetterNHot(".\lstm\midi")

.\lstm\midi
Found 8 composers: Bach, Beethoven, Chopin, Handel, Haydn, Hays, Thomas, Webster
Found 6432 training and 2145 test MIDI files!

C:\Users\livb1\Anaconda3\lib\site-packages\sklearn\preprocessing\_encoders.py:415: FutureWarning: The handling of integer data will change in version 0.22. Currently, the categories are determined based on the range [0, max(values)], while in the future they will be determined based on the unique values.
If you want the future behaviour and silence this warning, you can specify "categories='auto'".
In case you used a LabelEncoder before this OneHotEncoder to convert the categories to integers, then you can now use the OneHotEncoder directly.
  warnings.warn(msg, FutureWarning)
In [8]:
dataset.get_composers()

Found 8 composers: Bach, Beethoven, Chopin, Handel, Haydn, Hays, Thomas, Webster

C:\Users\livb1\Anaconda3\lib\site-packages\sklearn\preprocessing\_encoders.py:415: FutureWarning: The handling of integer data will change in version 0.22. Currently, the categories are determined based on the range [0, max(values)], while in the future they will be determined based on the unique values.
If you want the future behaviour and silence this warning, you can specify "categories='auto'".
In case you used a LabelEncoder before this OneHotEncoder to convert the categories to integers, then you can now use the OneHotEncoder directly.
  warnings.warn(msg, FutureWarning)
Out[8]:
array(['Bach', 'Beethoven', 'Chopin', 'Handel', 'Haydn', 'Hays', 'Thomas',
       'Webster'], dtype=object)