| 1 | #!/usr/bin/env python
|
|---|
| 2 |
|
|---|
| 3 | import csv
|
|---|
| 4 | import re
|
|---|
| 5 | from collections import OrderedDict
|
|---|
| 6 |
|
|---|
| 7 | def csv_features_to_dict(csv_filename):
|
|---|
| 8 | csv_features_file = open(csv_filename)
|
|---|
| 9 | csv_features_reader = csv.reader(csv_features_file, delimiter=',', quotechar='"')
|
|---|
| 10 |
|
|---|
| 11 | csv_features_dict = {}
|
|---|
| 12 |
|
|---|
| 13 | line_count = 0
|
|---|
| 14 | header_row = None
|
|---|
| 15 |
|
|---|
| 16 | for row in csv_features_reader:
|
|---|
| 17 | if line_count == 0:
|
|---|
| 18 | row.pop(0)
|
|---|
| 19 | header_row = row
|
|---|
| 20 | else:
|
|---|
| 21 | full_file_id = row[0]
|
|---|
| 22 |
|
|---|
| 23 | pat = re.compile('^import/(\\d+).json$')
|
|---|
| 24 | mat = pat.match(full_file_id)
|
|---|
| 25 | song_id = int(mat.group(1))
|
|---|
| 26 |
|
|---|
| 27 | row.pop(0)
|
|---|
| 28 | row_ordered_dict = OrderedDict()
|
|---|
| 29 |
|
|---|
| 30 | for i in range(0, len(header_row)):
|
|---|
| 31 | field = header_row[i];
|
|---|
| 32 | value = row[i];
|
|---|
| 33 |
|
|---|
| 34 | row_ordered_dict[field] = value
|
|---|
| 35 |
|
|---|
| 36 | csv_features_dict[song_id] = row_ordered_dict
|
|---|
| 37 |
|
|---|
| 38 | line_count += 1
|
|---|
| 39 |
|
|---|
| 40 | return csv_features_dict
|
|---|
| 41 |
|
|---|
| 42 | def csv_groundtruth_to_dict(csv_filename):
|
|---|
| 43 |
|
|---|
| 44 | csv_file = open(csv_filename)
|
|---|
| 45 | csv_reader = csv.reader(csv_file, delimiter=',')
|
|---|
| 46 |
|
|---|
| 47 | csv_gt_dict = {}
|
|---|
| 48 |
|
|---|
| 49 | line_count = 0
|
|---|
| 50 | header_row = None
|
|---|
| 51 |
|
|---|
| 52 | for row in csv_reader:
|
|---|
| 53 | if line_count == 0:
|
|---|
| 54 | row.pop(0)
|
|---|
| 55 | header_row = row
|
|---|
| 56 | else:
|
|---|
| 57 | song_id = int(row[0])
|
|---|
| 58 | row.pop(0)
|
|---|
| 59 |
|
|---|
| 60 | row_ordered_dict = OrderedDict()
|
|---|
| 61 |
|
|---|
| 62 | for i in range(0, len(row)):
|
|---|
| 63 | field = header_row[i];
|
|---|
| 64 | value = row[i];
|
|---|
| 65 |
|
|---|
| 66 | row_ordered_dict[field] = value
|
|---|
| 67 |
|
|---|
| 68 | csv_gt_dict[song_id] = row_ordered_dict
|
|---|
| 69 |
|
|---|
| 70 | line_count += 1
|
|---|
| 71 |
|
|---|
| 72 | return csv_gt_dict
|
|---|
| 73 |
|
|---|
| 74 |
|
|---|
| 75 | csv_features_dict = csv_features_to_dict('etc/deam-essentia-features-collated.csv')
|
|---|
| 76 |
|
|---|
| 77 | print("Essentia Features (sample of first 3 ids):")
|
|---|
| 78 | i = 1
|
|---|
| 79 | for song_id_key, ordered_feature_vals in sorted(csv_features_dict.items()):
|
|---|
| 80 | # print(str(song_id_key) + "\n" + str(ordered_feature_vals))
|
|---|
| 81 | if i >= 3:
|
|---|
| 82 | break
|
|---|
| 83 | i = i + 1
|
|---|
| 84 |
|
|---|
| 85 |
|
|---|
| 86 | groundtruth_dir = 'prepare/annotations/annotations averaged per song/dynamic (per second annotations)'
|
|---|
| 87 | arousal_csv_filename = groundtruth_dir + '/arousal.csv'
|
|---|
| 88 | valence_csv_filename = groundtruth_dir + '/valence.csv'
|
|---|
| 89 |
|
|---|
| 90 |
|
|---|
| 91 | arousal_groundtruth_dict = csv_groundtruth_to_dict(arousal_csv_filename)
|
|---|
| 92 | valence_groundtruth_dict = csv_groundtruth_to_dict(valence_csv_filename)
|
|---|
| 93 |
|
|---|
| 94 | print("Arousal Ground-truth (sample of first 3 ids):")
|
|---|
| 95 | i = 1
|
|---|
| 96 | for song_id_key, ordered_feature_vals in sorted(arousal_groundtruth_dict.items()):
|
|---|
| 97 | print(str(song_id_key) + "\n" + str(ordered_feature_vals))
|
|---|
| 98 | if i >= 3:
|
|---|
| 99 | break
|
|---|
| 100 | i = i + 1
|
|---|
