1
00:00:00,000 --> 00:00:03,955
All right, so I've got my motion detection

2
00:00:03,979 --> 00:00:07,320
program running here, and as

3
00:00:07,320 --> 00:00:09,660
you see this is capturing me moving at

4
00:00:09,660 --> 00:00:12,721
the moment. If I move away,

5
00:00:12,746 --> 00:00:15,151
[No audio]

6
00:00:15,194 --> 00:00:16,890
you can only see some shadows there,

7
00:00:16,915 --> 00:00:18,915
[No audio]

8
00:00:18,940 --> 00:00:20,640
and so this

9
00:00:20,640 --> 00:00:23,820
is great. But as I said, you may want to

10
00:00:23,820 --> 00:00:26,160
store the time, so when an object enters

11
00:00:26,435 --> 00:00:28,388
the frame, so that you end up

12
00:00:28,412 --> 00:00:29,621
with something like this.

13
00:00:29,645 --> 00:00:32,191
[No audio]

14
00:00:32,216 --> 00:00:36,030
So this is the first entrance, as the start time here

15
00:00:36,030 --> 00:00:38,160
is where the object exited the frame,

16
00:00:38,190 --> 00:00:41,430
and then the second entrance of the

17
00:00:41,430 --> 00:00:45,870
object, and so on. So in this

18
00:00:45,870 --> 00:00:48,180
lecture, I'll show you how to build

19
00:00:48,180 --> 00:00:51,150
that table using this script. So we

20
00:00:51,150 --> 00:00:55,260
will add some more code here, and the way to

21
00:00:55,260 --> 00:00:58,320
start doing this is to first figure out

22
00:00:58,680 --> 00:01:01,080
the point, the line in the script where

23
00:01:01,080 --> 00:01:04,470
the status is changing from, let's say,

24
00:01:04,715 --> 00:01:08,767
motion to non-motion. Hence, you, we have

25
00:01:08,798 --> 00:01:12,427
non-motion in here. So as soon as

26
00:01:12,606 --> 00:01:16,565
webcam triggers, let's say the status is

27
00:01:16,590 --> 00:01:20,160
0. So just a symbol that I'm choosing

28
00:01:20,160 --> 00:01:23,100
to denote that there's no motion in the

29
00:01:23,100 --> 00:01:26,100
current frame. So then you get that

30
00:01:26,370 --> 00:01:28,590
first_frame, which has a status of 0,

31
00:01:29,130 --> 00:01:33,270
and then you grayed out, and then you check

32
00:01:33,270 --> 00:01:34,740
if the first_frame is None and then you

33
00:01:34,740 --> 00:01:37,200
continue the loop again, and then you grab

34
00:01:37,200 --> 00:01:41,280
the second frame, so the status is still

35
00:01:41,280 --> 00:01:46,260
0, and then you apply the difference and the

36
00:01:46,260 --> 00:01:50,130
threshold, and the dilate, and then you iterate

37
00:01:50,130 --> 00:01:53,520
through contours. So if you find an area

38
00:01:53,520 --> 00:01:55,980
that is less than, I changed this

39
00:01:56,010 --> 00:02:00,150
earlier, I put it 10000. So I wanted to

40
00:02:00,150 --> 00:02:02,640
detect bigger objects, because I was

41
00:02:02,670 --> 00:02:04,590
closer to the camera, so I wanted to

42
00:02:04,590 --> 00:02:07,380
adjust it. So depending on what objects

43
00:02:07,380 --> 00:02:09,930
you want to capture, here you want to adjust

44
00:02:09,930 --> 00:02:10,971
these pixels here.

45
00:02:10,995 --> 00:02:12,995
[No audio]

46
00:02:13,024 --> 00:02:15,855
So 10000 is something like

47
00:02:16,112 --> 00:02:18,956
100 by 100 pixels window.

48
00:02:19,410 --> 00:02:22,350
So we're checking if you don't find a

49
00:02:22,350 --> 00:02:25,410
window that is bigger than 10000, then

50
00:02:25,410 --> 00:02:27,210
you continue to the beginning of the

51
00:02:27,210 --> 00:02:30,390
loop until you find a contour with

52
00:02:30,420 --> 00:02:33,450
this size bigger than this. When Python

53
00:02:33,450 --> 00:02:36,810
finds that contour with that size, we

54
00:02:36,810 --> 00:02:40,464
want to change the status to 1.

55
00:02:40,488 --> 00:02:42,912
[No audio]

56
00:02:42,937 --> 00:02:43,943
Okay.

57
00:02:45,230 --> 00:02:49,388
And let's now print out that status variable.

58
00:02:49,412 --> 00:02:51,681
[No audio]

59
00:02:51,706 --> 00:02:54,240
So what you want to do is

60
00:02:55,266 --> 00:02:59,376
print it at the end of your loop, so status.

61
00:03:00,900 --> 00:03:03,720
And let me delete this NumPy arrays,

62
00:03:03,750 --> 00:03:06,060
which are being printed, we don't need

63
00:03:06,060 --> 00:03:09,540
those, and make some room here.

64
00:03:09,564 --> 00:03:14,525
[No audio]

65
00:03:14,550 --> 00:03:18,780
So this is inside the loop, and I now expect

66
00:03:18,810 --> 00:03:21,600
to see the status, so 1 or 0 or 0 printed

67
00:03:21,625 --> 00:03:25,200
out depending on whether there is an

68
00:03:25,200 --> 00:03:28,230
object in the frame or not. So if this

69
00:03:28,230 --> 00:03:31,347
is working, that means we now have

70
00:03:31,574 --> 00:03:34,020
actual status of the frame, and we can

71
00:03:34,020 --> 00:03:37,688
use that 1 or 0 for other things. So let's see.

72
00:03:37,712 --> 00:03:42,027
[No audio]

73
00:03:42,052 --> 00:03:45,509
Okay, there is nothing at the moment. Let me

74
00:03:45,534 --> 00:03:47,534
[No audio]

75
00:03:47,559 --> 00:03:49,380
minimize this. So we can

76
00:03:49,380 --> 00:03:50,655
see the console.

77
00:03:50,679 --> 00:03:52,721
[No audio]

78
00:03:52,746 --> 00:03:55,860
This as well. Great. So

79
00:03:55,860 --> 00:03:58,110
0 is being printed out, and if

80
00:03:58,110 --> 00:04:00,780
something shows up, now you see 1s.

81
00:04:00,804 --> 00:04:04,161
[No audio]

82
00:04:04,186 --> 00:04:10,621
Again, 1, 0, 1, 1, 0, great.

83
00:04:10,645 --> 00:04:12,876
[No audio]

84
00:04:12,901 --> 00:04:17,790
We can now simply apply a datetime method,

85
00:04:18,180 --> 00:04:21,120
you know, such as datetime.now,

86
00:04:22,920 --> 00:04:25,590
and that would record the time of

87
00:04:25,590 --> 00:04:28,740
each frame. So we don't want that

88
00:04:28,740 --> 00:04:31,921
exactly. We want to use a datetime.now

89
00:04:31,945 --> 00:04:34,530
method but not like that. Because

90
00:04:34,530 --> 00:04:36,180
that would record times for

91
00:04:36,180 --> 00:04:40,020
every frame. So what we want instead is

92
00:04:40,800 --> 00:04:43,200
we want to know when the status changes

93
00:04:43,200 --> 00:04:46,530
from 0 to 1, which indicates that

94
00:04:46,560 --> 00:04:48,780
there is an object entering the frame.

95
00:04:49,110 --> 00:04:51,690
So we want to record that time, and we

96
00:04:51,690 --> 00:04:53,880
also want to record the other time when

97
00:04:54,450 --> 00:04:57,000
the status changes from 1 to 0, so

98
00:04:57,030 --> 00:04:59,940
the object exits the frame. Let's say

99
00:05:00,319 --> 00:05:03,379
status_list, that will be equal to

100
00:05:03,690 --> 00:05:06,570
empty brackets, so an empty list, and

101
00:05:06,570 --> 00:05:10,920
then I want to append the status of the

102
00:05:10,920 --> 00:05:12,807
object to that list.

103
00:05:14,497 --> 00:05:16,688
So after this loop finishes,

104
00:05:17,643 --> 00:05:22,020
which is here, so I want to be

105
00:05:22,020 --> 00:05:24,000
here at the same level with the loop, so I

106
00:05:24,025 --> 00:05:25,705
want to apply something outside the loop,

107
00:05:26,970 --> 00:05:31,950
so status_list, I want to append the

108
00:05:31,950 --> 00:05:34,470
current status of the list here in the

109
00:05:34,470 --> 00:05:39,180
status_list. Great, and to illustrate,

110
00:05:39,480 --> 00:05:41,970
to show you what this status_list is

111
00:05:41,970 --> 00:05:45,121
about, I'd like to print that out actually,

112
00:05:45,145 --> 00:05:48,383
[No audio]

113
00:05:48,408 --> 00:05:50,460
and I would want to print out

114
00:05:50,760 --> 00:05:53,130
a list outside the while loop. So I

115
00:05:53,130 --> 00:05:55,188
don't want to print out every

116
00:05:55,212 --> 00:05:56,788
time the loop iterates.

117
00:05:56,812 --> 00:06:00,817
[No audio]

118
00:06:00,842 --> 00:06:03,121
So no movement, movement here,

119
00:06:03,145 --> 00:06:07,572
[No audio]

120
00:06:07,597 --> 00:06:10,770
and quit. So these are the statuses

121
00:06:10,770 --> 00:06:14,340
for each of the frames, so 0, 0, 0, no movement,

122
00:06:14,370 --> 00:06:18,300
and then something entered the frame, and

123
00:06:18,300 --> 00:06:21,870
then something entered, like, let's

124
00:06:21,870 --> 00:06:25,985
say 30 frames and then the object exits

125
00:06:26,010 --> 00:06:30,750
the frame, and that's it. So what we

126
00:06:30,750 --> 00:06:33,570
want to do now is we want to capture, so

127
00:06:33,600 --> 00:06:36,240
what's next. So the next thing is we

128
00:06:36,240 --> 00:06:39,480
want to capture the time that we switch

129
00:06:39,480 --> 00:06:43,050
from 0 to 1, and also from 1 to

130
00:06:43,050 --> 00:06:45,720
0. So basically, what we're

131
00:06:45,720 --> 00:06:49,170
interested about is to capture the last

132
00:06:49,200 --> 00:06:53,940
two items on the status_list. So let's

133
00:06:53,940 --> 00:06:56,730
say we have status 0, so we

134
00:06:56,730 --> 00:06:59,400
start with a empty list, and the first item

135
00:06:59,400 --> 00:07:02,160
that will be appended to the status_list

136
00:07:02,225 --> 00:07:06,005
is 0, and then the second item, let's

137
00:07:06,030 --> 00:07:08,730
say will be 0, or let's say will be

138
00:07:08,730 --> 00:07:12,180
1, and so we want to apply a conditional

139
00:07:12,180 --> 00:07:15,360
to check whether the last two items of

140
00:07:15,360 --> 00:07:19,200
the lists are changing. So I know all this

141
00:07:19,230 --> 00:07:21,630
may be confusing, but once I write the,

142
00:07:21,660 --> 00:07:23,700
this four lines of code here, you'll

143
00:07:23,700 --> 00:07:25,800
understand it, and it will make a lot of

144
00:07:25,825 --> 00:07:30,813
sense. So we want to check if status_list,

145
00:07:30,941 --> 00:07:34,388
so the last item of the status_list,

146
00:07:34,985 --> 00:07:37,200
which has an index of -1,

147
00:07:37,230 --> 00:07:40,350
remember, list indexing, if that is

148
00:07:40,350 --> 00:07:41,616
equal to 1,

149
00:07:41,640 --> 00:07:43,898
[No audio]

150
00:07:43,923 --> 00:07:48,540
and status_list with index

151
00:07:48,565 --> 00:07:53,190
of -2, so the item that was added

152
00:07:53,190 --> 00:07:56,102
first, the item that was added second,

153
00:07:57,729 --> 00:07:58,957
is equal to 0.

154
00:07:58,981 --> 00:08:01,323
[No audio]

155
00:08:01,348 --> 00:08:03,240
We want to record the

156
00:08:03,240 --> 00:08:06,480
datetime of this event in the list. So

157
00:08:06,506 --> 00:08:09,206
let's create a list here, say times

158
00:08:10,593 --> 00:08:15,513
equals empty, and then we go here, here,

159
00:08:15,775 --> 00:08:22,554
and we say times.append datetime.now,

160
00:08:23,137 --> 00:08:24,540
you know about this function

161
00:08:24,540 --> 00:08:27,840
already, and hence so we want to append the

162
00:08:27,900 --> 00:08:30,894
current timestamp when this change occurred.

163
00:08:30,919 --> 00:08:32,927
[No audio]

164
00:08:32,952 --> 00:08:35,571
So that means we need to import

165
00:08:35,595 --> 00:08:38,310
[Author typing]

166
00:08:38,341 --> 00:08:42,246
or from datetime import

167
00:08:42,389 --> 00:08:48,220
datetime. So datetime class from the datetime library.

168
00:08:48,244 --> 00:08:50,944
[No audio]

169
00:08:50,969 --> 00:08:53,088
So we're basically recording the

170
00:08:53,117 --> 00:08:55,175
time when a status

171
00:08:55,200 --> 00:08:57,600
changed to 1, and we also wants to

172
00:08:57,600 --> 00:09:01,050
record the time when the status changes

173
00:09:01,050 --> 00:09:03,750
from 1 to 0. So we would have a

174
00:09:03,750 --> 00:09:05,460
similar conditional

175
00:09:06,690 --> 00:09:10,560
but this will be 0 and this would be

176
00:09:10,560 --> 00:09:13,170
1, okay.

177
00:09:13,194 --> 00:09:18,095
[No audio]

178
00:09:18,120 --> 00:09:21,990
So what's happening here is let's go

179
00:09:22,015 --> 00:09:25,465
back to the start of the loop. So,

180
00:09:25,805 --> 00:09:29,315
Python captured the first_frame and it

181
00:09:29,340 --> 00:09:32,670
converts it to gray and blurred it and

182
00:09:32,670 --> 00:09:36,221
apply GaussianBlur there to the first_frame,

183
00:09:36,245 --> 00:09:38,190
and then we store the first_frame

184
00:09:38,220 --> 00:09:40,170
in the first_frame variable and then we

185
00:09:40,170 --> 00:09:42,840
go ahead and calculate the difference

186
00:09:43,115 --> 00:09:44,760
between the first_frame and the current

187
00:09:44,786 --> 00:09:50,250
frame, findContours, and then if this

188
00:09:50,250 --> 00:09:52,440
little findContours greater than

189
00:09:52,440 --> 00:09:56,640
10000, a status variable changes from

190
00:09:56,640 --> 00:09:59,490
0, so it started at 0, so it changes

191
00:09:59,490 --> 00:10:03,150
at 1 here, and that means this empty

192
00:10:03,150 --> 00:10:06,600
status_list will get the another value. But

193
00:10:06,630 --> 00:10:09,720
the first value is always 0. So the

194
00:10:09,720 --> 00:10:13,380
list will start with 0, you know,

195
00:10:13,380 --> 00:10:16,288
like here, so the very first value is 0,

196
00:10:18,015 --> 00:10:20,550
and so on. Let's say for the next 10

197
00:10:20,550 --> 00:10:24,990
frames, it will be 0, 0, 0, 0, and so on, and

198
00:10:24,990 --> 00:10:28,560
then if an object enters the frame and then the

199
00:10:28,560 --> 00:10:30,810
status changes to 1, so we'll have

200
00:10:30,810 --> 00:10:34,920
0, 0, 0, 0, and then 1, and then what we're

201
00:10:34,920 --> 00:10:38,790
doing here is, we're comparing the last

202
00:10:38,872 --> 00:10:43,642
two items of the list, okay. And here, I'd like to

203
00:10:44,105 --> 00:10:48,335
print out the times.list, so that you see

204
00:10:48,360 --> 00:10:51,990
what's going on, and I expect to get an

205
00:10:52,015 --> 00:10:54,595
error at the first execution here,

206
00:10:54,930 --> 00:10:56,730
because we have something more to do.

207
00:10:56,730 --> 00:10:57,860
But let's see.

208
00:10:57,884 --> 00:11:02,055
[No audio]

209
00:11:02,080 --> 00:11:03,420
So yeah, it says list

210
00:11:03,420 --> 00:11:06,990
index is out of range. So that means

211
00:11:06,990 --> 00:11:10,350
Python is having troubles accessing the

212
00:11:10,380 --> 00:11:13,260
second item of the list, because the

213
00:11:13,260 --> 00:11:16,320
list has no items here, so it has no

214
00:11:16,320 --> 00:11:19,530
items, and then we add an item, so we

215
00:11:19,530 --> 00:11:22,950
add status, and then Python tries to

216
00:11:22,950 --> 00:11:24,510
find the second item, but there is no

217
00:11:24,510 --> 00:11:26,970
second item. So we need to do a trick

218
00:11:26,970 --> 00:11:31,830
here, we need to add two items, okay. Let's

219
00:11:31,830 --> 00:11:33,088
say None and None.

220
00:11:34,443 --> 00:11:36,030
Now let's execute the script again.

221
00:11:36,054 --> 00:11:41,225
[No audio]

222
00:11:41,262 --> 00:11:43,722
So there's no object there at the moment,

223
00:11:44,940 --> 00:11:48,330
now an object appears. It goes out

224
00:11:48,355 --> 00:11:54,870
again, two times, and one more time, so

225
00:11:54,870 --> 00:11:56,107
three entrances,

226
00:11:56,131 --> 00:11:58,640
[No audio]

227
00:11:58,665 --> 00:12:01,080
and I quit it. So let

228
00:12:01,080 --> 00:12:03,150
me go through this list now. This is

229
00:12:03,175 --> 00:12:07,585
our times.list, and so this is the time,

230
00:12:07,745 --> 00:12:12,488
so the year, the month, a day, and hour,

231
00:12:12,512 --> 00:12:15,270
minutes, and seconds, and then

232
00:12:15,270 --> 00:12:17,880
microseconds in here. So this is the

233
00:12:17,880 --> 00:12:21,210
start time when an object entered the

234
00:12:21,210 --> 00:12:23,640
frame for the first time, and then this

235
00:12:23,640 --> 00:12:25,590
is the time when the object exits the

236
00:12:25,590 --> 00:12:28,740
frame. So I stayed there for 2 seconds,

237
00:12:28,740 --> 00:12:34,170
so 51 upto 53, and then the next object

238
00:12:34,170 --> 00:12:39,600
enter the frame, so from this seconds to

239
00:12:39,625 --> 00:12:43,225
this, so for 2 other seconds, and then

240
00:12:43,260 --> 00:12:48,030
another third time, from 0 to 1, for

241
00:12:48,030 --> 00:12:52,105
1 second. So we've got three entrances

242
00:12:52,130 --> 00:12:55,500
in the frame. So I hope that makes

243
00:12:55,500 --> 00:12:58,800
sense. This is good. But sometimes, see

244
00:12:58,800 --> 00:13:02,121
it can happen that when you're running the script,

245
00:13:02,145 --> 00:13:04,338
[No audio]

246
00:13:04,363 --> 00:13:06,870
so you have background, and then

247
00:13:06,895 --> 00:13:10,675
the object shows, and the object disappears,

248
00:13:11,280 --> 00:13:14,520
and shows again, and then if you quit the

249
00:13:14,520 --> 00:13:17,010
program at this point, so let me quit it,

250
00:13:17,034 --> 00:13:19,034
[No audio]

251
00:13:19,059 --> 00:13:21,450
you'll see that you have the start time

252
00:13:21,725 --> 00:13:25,085
here for the first entrance, then the

253
00:13:25,110 --> 00:13:28,299
end time, and then you have another start

254
00:13:28,324 --> 00:13:29,880
time here for the second entrance.

255
00:13:30,155 --> 00:13:33,155
But then you don't have the exit time. So

256
00:13:33,186 --> 00:13:36,568
you may want to add the end time of

257
00:13:36,593 --> 00:13:39,210
the exit of your last object, which

258
00:13:39,210 --> 00:13:41,820
happens to be when the window was quit.

259
00:13:42,210 --> 00:13:45,840
So you want to go here at the end, and

260
00:13:45,865 --> 00:13:47,790
the script is waiting for a key, so if

261
00:13:47,790 --> 00:13:51,900
you press quit here, the while loop will

262
00:13:51,900 --> 00:13:54,510
break. But before that, you may want to

263
00:13:54,535 --> 00:13:57,235
check the status. So if status is 1,

264
00:13:57,777 --> 00:14:00,297
and in that case, you'd add knob with

265
00:14:00,330 --> 00:14:03,840
this scenario here. So if status is 1,

266
00:14:03,840 --> 00:14:06,521
when you press the Q key from the keyboard,

267
00:14:06,545 --> 00:14:08,647
[No audio]

268
00:14:08,672 --> 00:14:11,010
you want to add another item

269
00:14:11,035 --> 00:14:15,113
to your times.list, so times.append

270
00:14:16,480 --> 00:14:20,150
datetime.now. Great.

271
00:14:20,174 --> 00:14:22,345
[No audio]

272
00:14:22,370 --> 00:14:25,260
So that should work differently now. So

273
00:14:25,285 --> 00:14:26,288
let's execute the script.

274
00:14:26,312 --> 00:14:28,656
[No audio]

275
00:14:28,681 --> 00:14:33,725
No object, object, no object, object in the frame,

276
00:14:33,750 --> 00:14:36,930
quit, and now we should have four

277
00:14:36,930 --> 00:14:42,150
datetimes here, four timestamps 1, 2, 3, and

278
00:14:42,150 --> 00:14:47,940
4. So this here was recorded in this

279
00:14:47,975 --> 00:14:52,985
expression here. Great. And the next thing

280
00:14:53,010 --> 00:14:56,310
we want to do now is to throw this list

281
00:14:56,340 --> 00:14:59,160
into a Pandas DataFrame, and then into a

282
00:14:59,160 --> 00:15:01,500
csv file, so we'll have a start

283
00:15:01,500 --> 00:15:03,150
and an end column, and in the start

284
00:15:03,150 --> 00:15:05,520
column, we would have, for instance,

285
00:15:05,970 --> 00:15:09,060
this value, and in the end column, we

286
00:15:09,060 --> 00:15:11,280
would have the other values, so this one

287
00:15:11,280 --> 00:15:16,860
here, and the last one, this, okay, so first

288
00:15:16,860 --> 00:15:19,140
of all, we need an empty DataFrame, a

289
00:15:19,140 --> 00:15:21,270
Pandas DataFrame. So we'll be using

290
00:15:21,295 --> 00:15:24,565
Pandas to create DataFrames, of course.

291
00:15:25,429 --> 00:15:29,749
So let me import pandas here, and let's

292
00:15:29,814 --> 00:15:36,718
say, DataFrame, df equals to pandas.DataFrame,

293
00:15:36,743 --> 00:15:39,158
[No audio]

294
00:15:39,183 --> 00:15:40,252
with columns

295
00:15:40,276 --> 00:15:42,813
[No audio]

296
00:15:42,838 --> 00:15:47,880
Start and End, as easy as that. So that will

297
00:15:47,880 --> 00:15:50,400
create a DataFrame structure for us,

298
00:15:50,645 --> 00:15:54,064
which has no values, but it has two columns.

299
00:15:54,097 --> 00:15:56,097
[No audio]

300
00:15:56,128 --> 00:15:58,694
And then the next thing you want to do is

301
00:16:00,081 --> 00:16:01,950
outside the loop, so once

302
00:16:01,950 --> 00:16:03,450
you have created, once you have

303
00:16:03,450 --> 00:16:06,600
generated your times.list, you want to

304
00:16:06,600 --> 00:16:09,030
iterate through that list, and then

305
00:16:09,060 --> 00:16:11,430
append those values into the Pandas

306
00:16:11,430 --> 00:16:15,270
DataFrame. So you need a loop, let's

307
00:16:15,270 --> 00:16:21,960
say for i in range, and so for here,

308
00:16:21,960 --> 00:16:24,090
you'd have to iterate with a step of

309
00:16:24,120 --> 00:16:28,320
2 , and you'd want to iterate as many

310
00:16:28,320 --> 00:16:31,800
times as there are values in the list.

311
00:16:32,195 --> 00:16:36,185
So you'll start from 0, and then you'd

312
00:16:36,210 --> 00:16:39,030
end up at how many values you have in the

313
00:16:39,030 --> 00:16:40,770
list? Well, you can find that using the

314
00:16:40,770 --> 00:16:44,700
length method. So length times with a

315
00:16:44,700 --> 00:16:49,770
step of 2, great, so that will be, so

316
00:16:49,770 --> 00:16:52,500
for each iteration df, so the DataFrame

317
00:16:52,500 --> 00:16:57,330
will be equal to itself, .attend, so

318
00:16:57,330 --> 00:16:59,880
you are updating the DataFrame, and

319
00:16:59,880 --> 00:17:02,460
here you'd need to pass a dictionary. So

320
00:17:02,460 --> 00:17:06,330
for the Start column, you'd want to

321
00:17:06,360 --> 00:17:11,055
enter times i

322
00:17:11,916 --> 00:17:13,499
and for the End column,

323
00:17:13,530 --> 00:17:15,575
[No audio]

324
00:17:15,600 --> 00:17:18,420
so the keys of the dictionary here are

325
00:17:18,445 --> 00:17:21,655
Start and End, and you'd want to append

326
00:17:21,720 --> 00:17:25,288
times i+1,

327
00:17:25,992 --> 00:17:29,625
and you want to ignore_index,

328
00:17:31,116 --> 00:17:36,000
so equal to True. Great. So what

329
00:17:36,000 --> 00:17:39,180
this is doing is, let's say for i in

330
00:17:39,180 --> 00:17:41,520
range, this will be, let's say, we have a

331
00:17:41,520 --> 00:17:44,820
list with 6 items. So we have 3

332
00:17:44,820 --> 00:17:48,330
Start times and 3 End times. So that

333
00:17:48,330 --> 00:17:51,600
would mean range 0 to 6 with a

334
00:17:51,600 --> 00:17:54,660
step of 2. So for 0, we would do

335
00:17:54,660 --> 00:17:57,570
this. So we will start from 0 in the

336
00:17:57,570 --> 00:18:01,320
loop, and we would access the item with

337
00:18:01,320 --> 00:18:04,440
the 0 index in times, which happens

338
00:18:04,440 --> 00:18:06,951
to be this item for instance, so the first item

339
00:18:06,979 --> 00:18:08,979
[No audio]

340
00:18:09,004 --> 00:18:10,410
and then we append that item

341
00:18:10,410 --> 00:18:13,050
to the first row of the Start column.

342
00:18:13,074 --> 00:18:15,395
[No audio]

343
00:18:15,420 --> 00:18:18,255
And then we access the item 0+1

344
00:18:18,279 --> 00:18:20,040
which happens to be item with index

345
00:18:20,040 --> 00:18:22,980
1, which should be this item, so the

346
00:18:22,980 --> 00:18:25,560
second item, and we append that item to

347
00:18:25,560 --> 00:18:29,645
the End column for the same row. So

348
00:18:29,670 --> 00:18:31,890
that's it, then we go to the next item.

349
00:18:32,513 --> 00:18:36,653
So for so this time, we would start with

350
00:18:36,690 --> 00:18:42,270
index 3, so 3 here added to the

351
00:18:42,270 --> 00:18:44,760
Start column and then 4 added to the

352
00:18:44,760 --> 00:18:47,410
Start column to the add column, so we add

353
00:18:47,435 --> 00:18:50,490
so on, and lastly, we would have a

354
00:18:50,490 --> 00:18:53,580
DataFrame, and what you might want to

355
00:18:53,580 --> 00:18:56,688
do then is df.to_csv,

356
00:18:56,712 --> 00:18:59,672
[Author typing]

357
00:18:59,697 --> 00:19:01,749
let's say times.csv.

358
00:19:02,012 --> 00:19:04,088
So we are exporting the DataFrame

359
00:19:04,112 --> 00:19:05,588
to a csv file.

360
00:19:06,223 --> 00:19:08,700
Great. So let's see how this will go.

361
00:19:08,724 --> 00:19:15,515
[No audio]

362
00:19:15,540 --> 00:19:21,570
It's getting quite dark here. Okay, and

363
00:19:21,595 --> 00:19:25,318
one more time, frame, quit,

364
00:19:25,342 --> 00:19:27,886
[No audio]

365
00:19:27,911 --> 00:19:30,188
and as you see times.csv was

366
00:19:30,212 --> 00:19:32,610
created in my file system.

367
00:19:32,634 --> 00:19:35,315
[No audio]

368
00:19:35,340 --> 00:19:37,088
So these are the times.

369
00:19:37,125 --> 00:19:39,245
[No audio]

370
00:19:39,270 --> 00:19:41,620
I can also open it using Excel.

371
00:19:41,644 --> 00:19:45,385
[No audio]

372
00:19:45,410 --> 00:19:46,860
So these are the

373
00:19:46,860 --> 00:19:49,410
Start and End times. if you want to get

374
00:19:49,410 --> 00:19:52,290
four month then just four months sales,

375
00:19:53,070 --> 00:19:55,860
here and you can see the dates now.

376
00:19:56,070 --> 00:19:58,355
Great. Great.

377
00:19:58,379 --> 00:20:02,034
[No audio]

378
00:20:02,059 --> 00:20:04,505
So that ends this lecture.

379
00:20:04,530 --> 00:20:07,830
And I hope you enjoyed it. We have quite

380
00:20:07,830 --> 00:20:10,740
a good script now, and to make it more

381
00:20:10,770 --> 00:20:14,130
beautiful, we would want to display

382
00:20:14,160 --> 00:20:17,100
these times in a graph. So that's how

383
00:20:17,100 --> 00:20:19,290
you turn your data into real

384
00:20:19,290 --> 00:20:22,140
information. And we'll do that through the

385
00:20:22,140 --> 00:20:24,390
next lectures. But before I'd like to

386
00:20:24,420 --> 00:20:26,700
introduce you to the bokeh library,

387
00:20:26,945 --> 00:20:29,105
which is a library for interactive

388
00:20:29,130 --> 00:20:31,530
visualizations in the browser, so

389
00:20:31,530 --> 00:20:33,030
please follow the lectures in the

390
00:20:33,030 --> 00:20:35,880
correct order, and I'm sure you will

391
00:20:35,880 --> 00:20:38,373
learn a lot. Talk to you later.