core.py 13.8 KB
Newer Older
ecofractal's avatar
ecofractal committed
1
2
3
4
5
6
7
8
9
10
11
12
13
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Using specific functions for sonification. This code is partially inspired by
the project Sonify (https://github.com/erinspace/sonify) with some
modifications to be applied on environmental data.
Although some parameters were changed, we decided to keep some functions name
as originally defined by Erin Braswell at her work.
Currently our focus will be the temperature, conductivity and salinity data.
Further steps: biodiversity, birds and big mammals migratory data.
For visual analysis of discrete nonlinear dynamical systems most of the code
bases its function on Pynamical package.

hackecology's avatar
hackecology committed
14
TODO
ecofractal's avatar
ecofractal committed
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
Lacunarity Test - Measure of the nonuniformity (heterogeneity) of structure or
the degree of structural variance within an object

version 0.0.2
June, 7th 2019
sjacques 
"""


import csv
import io
import numpy as np
import pygame

from midiutil.MidiFile import MIDIFile
from pretty_midi import note_name_to_number
from time import sleep
from .midisource import KEYS, INSTRUMENTS, PERCUSSION



'''
FRACTAL GEOMETRY ANALYSIS
=========================

###########
#Box Count#
###########

Estimation of Fractal Dimension:
Minkowski–Bouligand dimension (computed)
Haussdorf dimension (theoretical)

'''
def rgb2gray(rgb):
    r, g, b = rgb[:,:,0], rgb[:,:,1], rgb[:,:,2]
    gray = 0.2989 * r + 0.5870 * g + 0.1140 * b
    return gray

'''
the threshold must to be set up according to the image
'''
def fracdim(Z, threshold):
    assert(len(Z.shape) == 2)
    
    def boxcount(Z, k):
        S = np.add.reduceat(
            np.add.reduceat(Z, np.arange(0, Z.shape[0], k), axis=0),
                               np.arange(0, Z.shape[1], k), axis=1)

        return len(np.where((S > 0) & (S < k*k))[0])

    Z = (Z < threshold)
    
    p = min(Z.shape)
    
    n = 2**np.floor(np.log(p)/np.log(2))

    n = int(np.log(n)/np.log(2))
    
    sizes = 2**np.arange(n, 1, -1)
    
    counts = []
    for size in sizes:
        counts.append(boxcount(Z, size))
        
    coeffs = np.polyfit(np.log(sizes), np.log(counts), 1)
    return -coeffs[0]



'''
SONIFICATION
============
'''

'''
92
Define the starting note based on the subtraction of y_values by notes_in_key
ecofractal's avatar
ecofractal committed
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
new_y.append create values based on the sum of y and traspose_value
'''
def make_first_number_match_key(y_values, notes_in_key):
    first_note_in_key = notes_in_key[0]
    transpose_value = first_note_in_key - y_values[0]
    new_y = []
    for y in y_values:
        new_y.append(y + transpose_value)

    return new_y

'''
    Convert a key name to notes, using C3=60
    :param key: String matching one of the values in pre-defined KEY dict
    :param octave_start: octave for the first note, as defined by C3=60
    :param number_of_octaves: The number of octaves to include in the list
'''
def key_name_to_notes(key, octave_start=1, number_of_octaves=4):
    key = KEYS.get(key)
    if not key:
        raise ValueError('Opps! No key by that name found')

    notes = []
    octave = octave_start + 1
    while len(notes) < number_of_octaves * 7:
        for note in key:
            note_with_octave = note + str(octave)
            note_number = note_name_to_number(note_with_octave)
            if note_number % 12 == 0 and len(notes) != 0:
                octave += 1
                note_with_octave = note + str(octave)
                note_number = note_name_to_number(note_with_octave)
            notes.append(note_number)

    return notes

'''
Define notes using the more close/possible value of MIDI.
It's done sorting the "possible values" according to lambda i: abs(i - value)
'''

def get_closest_midi_value(value, possible_values):
    return sorted(possible_values, key=lambda i: abs(i - value))[0]


'''
 midi notes have a range of 0 - 127. Make sure the data is in that range
    data: list of tuples of x, y coordinates for pitch and timing
    min: min data value, defaults to 0
    max: max data value, defaults to 127
    return: data, but y normalized to the range specified by min and max
'''
def scale_y_to_midi_range(data, new_min=0, new_max=127):
    if new_min < 0 or new_max > 127:
        raise ValueError('Midi notes must be in a range from 0 - 127')

    x, y = zip(*data)
    new_y = scale_list_to_range(y, new_min, new_max)

    return list(zip(x, new_y))


'''
Use the min and max values from "old_value" and the new parameters to set scaled
values inside the range
'''
def get_scaled_value(old_value, old_min, old_max, new_min, new_max):
    return ((old_value - old_min)/(old_max - old_min)) * (new_max - new_min) 
    + new_min


'''
165
Set a list inside a MIDI range defined by new_min and new_max
ecofractal's avatar
ecofractal committed
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
the output is based on the list_to_scale and uses get_scaled_value function
'''
def scale_list_to_range(list_to_scale, new_min, new_max):
    old_min = min(list_to_scale)
    old_max = max(list_to_scale)
    return [get_scaled_value(value, old_min, old_max, new_min, new_max) for 
            value in list_to_scale]


'''
Restrict the x range to something that's a multiple of the number of steps given
'''
def quantize_x_value(list_to_quantize, steps=0.5):
    quantized_x = []
    for x in list_to_quantize:
        quantized_x.append(round(steps * round(float(x) / steps), 2))
    return quantized_x

'''
Access the midisource.py file and set up the instrument sound
'''
def get_instrument(instrument_name):
    instrument_type = 'melodic'
    program_number = INSTRUMENTS.get(instrument_name.lower())
    if not program_number:
        program_number = PERCUSSION.get(instrument_name.lower())
        instrument_type = 'percussion'
        if not program_number:
            raise AttributeError('No instrument could be found by that name')
    return program_number - 1, instrument_type


'''
Define the key. As defined before, the percussion is a default. Otherwise it
will be necessary to use key_name_to_notes, make_first_number_match_key and
scale_list to_range to fit the range for other instruments.
line 140: Finding the index of the note closest to all the notes in the
options list
'''
def convert_to_key(data, key, number_of_octaves=4):
    instrument, instrument_type = None, None
    if type(data[0]) != tuple:
        instrument = data.pop(0)
        _, instrument_type = get_instrument(instrument)

    x, y = zip(*data)

    if instrument_type == 'percussion':
        new_y = y
    else:
        notes_in_key = key_name_to_notes(key, 
                                         number_of_octaves=number_of_octaves)

        transposed_y = make_first_number_match_key(y, notes_in_key)
        scaled_y = scale_list_to_range(transposed_y, new_min=min(notes_in_key), 
                                       new_max=max(notes_in_key))

        new_y = []
        for note in scaled_y:
            new_y.append(get_closest_midi_value(note, notes_in_key))

    processed_data = list(zip(x, new_y))

    if instrument:
        processed_data = [instrument] + processed_data

    return processed_data


'''
This function is applied to JSON data (in this case for climate data)
237
238
The dataset can be found here: https://www.ncdc.noaa.gov/cag/global/time-series
You can dowload the JSON file or just write the link
ecofractal's avatar
ecofractal committed
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
'''
def normalize_climate_data(climate_json):
    years = [int(year) for year in climate_json['data'].keys()]
    temp_anomolies = [float(temp_anomaly) for temp_anomaly in 
                      climate_json['data'].values()]

    normalized_years = scale_list_to_range(years, new_min=0, new_max=30)
    normalized_temp_anomolies = scale_list_to_range(temp_anomolies, new_min=30,
                                                    new_max=127)


    normed_climate_data = list(zip(normalized_years, normalized_temp_anomolies))
    
    return normed_climate_data


'''
Import df to list and certify we don't have NaN
With this code we generate a nested list of tuples for multitrack analisys
'''

def normalize_climate_multi(df):
    df = df.replace(np.nan, 0)

#From df to list: year is the key and the othe variables will be its values
    years_list = [int(year) for year in df['Date'].keys()]
    temperature_list = [float(temp) for temp in df['Temperature'].tolist()]
    conductivity_list = [float(conduct) for conduct in df['Conduct'].tolist()]
    salinity_list = [float(sal) for sal in df['Salinity'].tolist()]
    
#normalize data
    normalized_years_multi = scale_list_to_range(years_list, new_min=0, 
                                                 new_max=30)
    normalized_temp_multi = scale_list_to_range(temperature_list, new_min=30, 
                                                new_max=127)
    normalized_cond_multi = scale_list_to_range(conductivity_list, new_min=30, 
                                                new_max=127)
    normalized_sal_multi = scale_list_to_range(salinity_list, new_min=30, 
                                               new_max=127)
    
    normed_climate_multi = list(zip( normalized_years_multi, 
                                    normalized_temp_multi))
    normed_cond_multi = list(zip( normalized_years_multi, 
                                 normalized_cond_multi))
    normed_sal_multi = list(zip( normalized_years_multi, 
                                normalized_sal_multi))
    
    normed_multi = [normed_climate_multi]+[normed_cond_multi]+[normed_sal_multi]
    
    return(normed_multi)


'''
To use JSON with MIDItime library. It converts the df to a dictionary

'''
def csv_to_MIDITime_data(filename):
    mydata = []
    with open(filename, 'r') as f:
        reader=csv.reader(f)
        next(reader, None) #this is added only in case of a file with header
        for row in reader:
            mydict = {'days_since_epoch': float(row[0]) , 
                      'magnitude': float(row[1])}
            mydata.append(mydict)
    return mydata


"""
Export the MIDIfile
data: dictionary of x, y coordinates for pitch and timing
Optional: add a string to the start of the data list to specify instrument!
type: the type of data passed to create tracks. Either 'single' or 'multiple'
312
313
The bpm is defined as 120 and might be changed according to the input data.

ecofractal's avatar
ecofractal committed
314
315
"""

316
def write_to_midifile(data, bpm = 120, track_type='single'):
ecofractal's avatar
ecofractal committed
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335

    if track_type not in ['single', 'multiple']:
        raise ValueError('Track type must be single or multiple')

    if track_type == 'single':
        data = [data]

    memfile = io.BytesIO()
    midifile = MIDIFile(numTracks=len(data), adjust_origin=False)

    track = 0
    time = 0
    program = 0
    channel = 0
    duration = 1
    volume = 90

    for data_list in data:
        midifile.addTrackName(track, time, 'Track {}'.format(track))
336
        midifile.addTempo(track, time, bpm)
ecofractal's avatar
ecofractal committed
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360

        instrument_type = 'melodic'
        if type(data_list[0]) != tuple:
            program, instrument_type = get_instrument(data_list.pop(0))

        if instrument_type == 'percussion':
            volume = 100
            channel = 9

        # Write the notes we want to appear in the file
        for point in data_list:
            time = point[0]
            pitch = int(point[1]) if instrument_type == 'melodic' else program
            midifile.addNote(track, channel, pitch, time, duration, volume)
            midifile.addProgramChange(track, channel, time, program)

        track += 1
        channel = 0

    midifile.writeFile(memfile)

    return memfile


361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
'''
Export MIDI file using 

'''
def export_midi(data, key=None, number_of_octaves=4, bpm = 120,
                track_type='single'):

    if track_type not in ['single', 'multiple']:
        raise ValueError('Track type must be single or multiple')

    if track_type == 'single':
        data = [data]

    expfile = io.BytesIO()
    midifile = MIDIFile(numTracks=len(data), adjust_origin=False)

    track = 0
    time = 0
    program = 0
    channel = 0
    duration = 1
    volume = 90

    for data_list in data:
        midifile.addTrackName(track, time, 'Track {}'.format(track))
        midifile.addTempo(track, time, bpm)

        instrument_type = 'melodic'
        if type(data_list[0]) != tuple:
            program, instrument_type = get_instrument(data_list.pop(0))

        if instrument_type == 'percussion':
            volume = 100
            channel = 9

        # Write the notes we want to appear in the file
        for point in data_list:
            time = point[0]
            pitch = int(point[1]) if instrument_type == 'melodic' else program
            midifile.addNote(track, channel, pitch, time, duration, volume)
            midifile.addProgramChange(track, channel, time, program)

        track += 1
        channel = 0

    expfile = open("output.mid", 'wb') 
    midifile.writeFile(expfile)
    midifile.close() 
    
    return expfile

ecofractal's avatar
ecofractal committed
412
413
'''
To play MIDI without having to save to a file
414
This is making use of pygame as we can see.
ecofractal's avatar
ecofractal committed
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
'''
def play_memfile_as_midi(memfile):
    pygame.init()
    pygame.mixer.init()
    memfile.seek(0)
    pygame.mixer.music.load(memfile)
    pygame.mixer.music.play()
    while pygame.mixer.music.get_busy():
        sleep(1)
    print('Done playing!')


"""
As input_data it is used a list of tuples, or a list of lists of tuples to add
as separate tracks

e.g:
    input_data = [(1, 7), (7, 9)]
        OR
    input_data = [
        [(1, 3), (5, 2)],
        [(4, 1), (3, 12)]
    ]

key: key to play back the graph -- see constants.py for current choices
number_of_octaves: number of octaves used to restrict the music playback
when converting to a key
optional -- append an instrument name to the start of each data list to play
back using that program number!
"""
def play_midi_from_data(input_data, key=None, number_of_octaves=4, 
                        track_type='single'):

    if key:
        if track_type == 'multiple':
            data = []
            for data_list in input_data:
                data.append(convert_to_key(data_list, key, number_of_octaves))
        else:
            data = convert_to_key(input_data, key, number_of_octaves)
    else:
        data = input_data

    memfile = write_to_midifile(data, track_type)
459
460
    play_memfile_as_midi(memfile)