- In this quick and dirty
DBSCAN, I take my iphone screenshots dropbox folder and use DBSCAN with600 seconds. - Output looks pretty accurate when I inspected the files themselves.
import re
import argparse
import os
import datetime
from sklearn.cluster import DBSCAN
from sklearn.preprocessing import StandardScaler
import random
def make_clusters(files_vec):
screenshots = sorted(filter(is_screenshot, files_vec))
dates = map(to_date, screenshots)
first_date = dates[0]
deltas = [(d - first_date).total_seconds() for d in dates]
db = DBSCAN(eps=600, min_samples=10).fit([[x] for x in deltas])
return zip(screenshots, deltas, db.labels_)
def is_screenshot(x):
return True if re.match('\d{4}-\d{2}-\d{2} \d{2}.\d{2}.\d{2}.png', x) else False
def to_date(x):
return datetime.datetime.strptime(x, '%Y-%m-%d %H.%M.%S.png')In [13]: out = make_clusters(os.listdir('.'))
In [14]: len(out)
Out[14]: 3669
In [15]: import pandas as pd
In [21]: df = pd.DataFrame.from_records(out, columns=['file', 'delta', 'label'])
In [22]: df.to_csv('/Users/me/Downloads/2019-01-13--dbscan--folder-clusters/output.csv')
In [22]: df[df.label != -1].shape
Out[22]: (899, 3)
In [23]: df[df.label != -1].head()
Out[23]:
file delta label
328 2018-01-27 12.27.44.png 184777822.0 0
329 2018-01-27 12.28.06.png 184777844.0 0
330 2018-01-27 12.28.47.png 184777885.0 0
331 2018-01-27 12.29.46.png 184777944.0 0
332 2018-01-27 12.30.41.png 184777999.0 0
In [24]: df[df.label != -1].label.value_counts()
Out[24]:
7 41
8 37
34 35
12 27
32 26
25 26
28 25
11 22
9 21
13 19
53 18
16 18
43 18
1 18
26 18
22 17
41 17
15 16
17 16
0 16
48 16
23 15
2 15
31 15
30 15
20 15
52 15
10 15
27 13
33 13
46 12
14 12
47 12
36 12
49 12
44 12
6 12
54 12
45 12
18 12
35 12
38 12
19 12
5 11
21 11
42 11
37 11
40 11
3 11
4 10
51 10
29 10
55 10
50 10
39 10
24 10
56 9
Name: label, dtype: int64
In [27]: print df[df.label != -1].label.unique().tolist()
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 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]
In [28]: df[df.label == 0]
Out[28]:
file delta label
328 2018-01-27 12.27.44.png 184777822.0 0
329 2018-01-27 12.28.06.png 184777844.0 0
330 2018-01-27 12.28.47.png 184777885.0 0
331 2018-01-27 12.29.46.png 184777944.0 0
332 2018-01-27 12.30.41.png 184777999.0 0
333 2018-01-27 12.32.11.png 184778089.0 0
334 2018-01-27 12.37.30.png 184778408.0 0
335 2018-01-27 12.37.54.png 184778432.0 0
336 2018-01-27 12.39.01.png 184778499.0 0
337 2018-01-27 12.41.24.png 184778642.0 0
338 2018-01-27 12.41.53.png 184778671.0 0
339 2018-01-27 12.42.32.png 184778710.0 0
340 2018-01-27 12.42.55.png 184778733.0 0
341 2018-01-27 12.43.26.png 184778764.0 0
342 2018-01-27 12.44.21.png 184778819.0 0
343 2018-01-27 12.46.28.png 184778946.0 0
In [29]: df[df.label == 1]
Out[29]:
file delta label
379 2018-02-05 16.34.24.png 185570222.0 1
380 2018-02-05 16.38.33.png 185570471.0 1
381 2018-02-05 16.38.58.png 185570496.0 1
382 2018-02-05 16.39.21.png 185570519.0 1
383 2018-02-05 16.39.26.png 185570524.0 1
384 2018-02-05 16.40.08.png 185570566.0 1
385 2018-02-05 16.42.44.png 185570722.0 1
386 2018-02-05 16.43.30.png 185570768.0 1
387 2018-02-05 16.44.09.png 185570807.0 1
388 2018-02-05 16.44.30.png 185570828.0 1
389 2018-02-05 16.45.43.png 185570901.0 1
390 2018-02-05 16.46.03.png 185570921.0 1
391 2018-02-05 16.47.45.png 185571023.0 1
392 2018-02-05 16.48.39.png 185571077.0 1
393 2018-02-05 16.48.51.png 185571089.0 1
394 2018-02-05 16.49.38.png 185571136.0 1
395 2018-02-05 16.50.58.png 185571216.0 1
396 2018-02-05 16.51.29.png 185571247.0 1
In [30]: df[df.label == 2]
Out[30]:
file delta label
403 2018-02-08 16.33.57.png 185829395.0 2
404 2018-02-08 16.34.30.png 185829428.0 2
405 2018-02-08 16.35.03.png 185829461.0 2
406 2018-02-08 16.35.08.png 185829466.0 2
407 2018-02-08 16.37.28.png 185829606.0 2
408 2018-02-08 16.41.50.png 185829868.0 2
409 2018-02-08 16.46.11.png 185830129.0 2
410 2018-02-08 16.46.52.png 185830170.0 2
411 2018-02-08 16.47.39.png 185830217.0 2
412 2018-02-08 16.48.25.png 185830263.0 2
413 2018-02-08 16.49.17.png 185830315.0 2
414 2018-02-08 16.50.21.png 185830379.0 2
415 2018-02-08 16.52.04.png 185830482.0 2
416 2018-02-08 16.52.31.png 185830509.0 2
417 2018-02-08 16.53.42.png 185830580.0 2