1
00:00:00,150 --> 00:00:06,480
Helen, before going deep dive into the session, let's have a quick recap of what we have done in the

2
00:00:06,480 --> 00:00:07,280
previous session.

3
00:00:07,620 --> 00:00:09,830
So basically we have imported our data.

4
00:00:09,840 --> 00:00:15,990
We have done lots of data processing on our data as well as we have deal with the missing values.

5
00:00:16,200 --> 00:00:20,760
We have simply dropped all the missing value because we have less missing values and after it we have

6
00:00:20,760 --> 00:00:23,600
to define its function to change my data type.

7
00:00:23,700 --> 00:00:28,740
Whatever the type of variable will be, it would just convert into some data format after doing all

8
00:00:28,740 --> 00:00:32,820
these things we have accepted as Journey Day journeyman for our column.

9
00:00:32,820 --> 00:00:36,650
And we have basically dropped this column because it makes no sense at all.

10
00:00:36,960 --> 00:00:43,140
And in the session, we have to deal with this arrival time as well as this departure time feature as

11
00:00:43,140 --> 00:00:47,550
well, because whenever you are going to parse this feature to a machine learning model, your model

12
00:00:47,550 --> 00:00:54,510
isn't able to understand what exactly that particular day, what exactly the particular arrival time,

13
00:00:54,780 --> 00:00:56,850
because you have to tell with machine learning model.

14
00:00:56,850 --> 00:00:57,190
Yeah.

15
00:00:57,510 --> 00:01:04,530
And I would offer this much my arrival, this much my departure hours did much, my departure this much.

16
00:01:04,770 --> 00:01:10,110
And after what we have to do for this, for performing all these things, what do we have to do?

17
00:01:10,110 --> 00:01:15,740
We have to fetch May not an hour from my data and we have to send to that machine learning model.

18
00:01:15,780 --> 00:01:17,120
That's what I want to do.

19
00:01:17,550 --> 00:01:25,240
So to extract work from just data to extract out from this column, I would define a function X.

20
00:01:25,290 --> 00:01:34,250
Its name is Lex, its name is extract, underscore our and on what data stream I have to apply that.

21
00:01:34,260 --> 00:01:36,870
It was at the very first parameter.

22
00:01:37,140 --> 00:01:40,430
The second parameter is on what column I have to apply.

23
00:01:41,100 --> 00:01:42,750
So now what I have to do.

24
00:01:42,930 --> 00:01:47,650
So basically I'm going to say the fourth column dot dot dot hour.

25
00:01:47,700 --> 00:01:50,350
So it will exactly return me some hour.

26
00:01:50,460 --> 00:01:52,560
So now I have to store it somewhere else.

27
00:01:52,810 --> 00:01:58,280
Let's say I have to store it in whatever column I want to pass in my function name.

28
00:01:58,290 --> 00:02:02,080
You have to just concatenate underscore our order.

29
00:02:02,280 --> 00:02:04,770
Similarly, you have to write a function.

30
00:02:04,770 --> 00:02:09,660
Similarly, you have to write a function to extract your minute as well.

31
00:02:09,670 --> 00:02:12,370
So I'm just going to write a function over there.

32
00:02:12,390 --> 00:02:17,460
So for this I have to assign proper indentation, otherwise it will give me some error.

33
00:02:17,700 --> 00:02:20,790
So this time I have to say this time I function.

34
00:02:20,790 --> 00:02:23,700
Name is say accoustic underscoring.

35
00:02:23,730 --> 00:02:31,320
And here I have to say I have to extract this minute and here I'm going to say column name is Will be

36
00:02:31,320 --> 00:02:31,890
Nothing.

37
00:02:31,890 --> 00:02:33,510
My column on this column.

38
00:02:33,870 --> 00:02:39,590
And just as an all this team likes it, after doing all this thing, you have to drop that column because

39
00:02:39,610 --> 00:02:41,850
that column makes no sense at all.

40
00:02:42,070 --> 00:02:45,000
So I will define mine, let's say some new function.

41
00:02:45,210 --> 00:02:50,910
Let's say in my function, name is nothing but drop on a score column to make it more user friendly.

42
00:02:50,910 --> 00:02:55,170
And after it, I have to pass two parameters, my dear friend column.

43
00:02:55,380 --> 00:02:57,640
Then I have to say what I have to draw.

44
00:02:57,720 --> 00:02:59,580
So I'm going to say, do you have to draw?

45
00:02:59,610 --> 00:03:02,550
And this time I have to pass my column.

46
00:03:02,550 --> 00:03:08,430
Then X is parameter as well as I have to pass my in-place parameter as well, because I have to admit

47
00:03:09,000 --> 00:03:09,470
as well.

48
00:03:09,750 --> 00:03:14,130
And after doing all this, what I have to do, I have to just executed.

49
00:03:14,130 --> 00:03:16,800
And let's say very first what I'm going to do.

50
00:03:16,920 --> 00:03:22,070
I have to apply all of these three functions on my this departure time.

51
00:03:22,350 --> 00:03:27,230
So let's say the very first I would apply all these functions on my departure.

52
00:03:27,240 --> 00:03:33,480
Let's go down so that if I have to say what function I have to play, so let's say I have to extract

53
00:03:33,510 --> 00:03:37,620
all those thoughts, I have to call this on a sort of assumption.

54
00:03:37,800 --> 00:03:43,410
And here I'm going to say my datastream name is nothing but my training data.

55
00:03:43,410 --> 00:03:49,450
And this time my column name is nothing or be on this call time.

56
00:03:49,490 --> 00:03:53,170
Make sure we don't have any case sensitive issues, otherwise it will give you error.

57
00:03:53,430 --> 00:03:58,290
Similarly, if you have to accept mynor, you guys can call this function as well.

58
00:03:58,500 --> 00:04:03,660
And here very first you have to Passo data from the second one feature you have to parse.

59
00:04:03,720 --> 00:04:05,970
What exactly is your column name.

60
00:04:06,120 --> 00:04:10,800
So I'm going to pass all this stuff and after doing all these things you have to drop that particular

61
00:04:10,800 --> 00:04:11,190
column.

62
00:04:11,490 --> 00:04:16,460
So I'm going to say my function name is nothing but drop on this column and my data framing the thing,

63
00:04:16,480 --> 00:04:22,170
what it's called data mining feature, multimode departure on this school time.

64
00:04:22,350 --> 00:04:25,350
And after it, we have to just execute it.

65
00:04:25,350 --> 00:04:31,950
And if this time I'm with the caller had already, we will see what we don't have any departure time

66
00:04:31,950 --> 00:04:32,460
column.

67
00:04:32,670 --> 00:04:39,690
And you have a two column names additionally over here as departure time and our deposit that on the

68
00:04:39,690 --> 00:04:40,250
school minutes.

69
00:04:40,560 --> 00:04:42,240
So that's a power function.

70
00:04:42,240 --> 00:04:48,660
Whenever you have to do a task multiple times, just like that block of code and you function whenever

71
00:04:48,660 --> 00:04:52,020
you have a need of that function, just call it that simple.

72
00:04:52,170 --> 00:04:58,650
Similarly, you have to perform all these stuffs for your arrival time feature as well, which is nothing.

73
00:04:58,650 --> 00:04:59,940
But which is this one.

74
00:05:00,210 --> 00:05:07,320
So I'm just going to copy this and just I'm going to paste over here this time here, I'm going to say

75
00:05:07,620 --> 00:05:10,950
my future is nothing but arrival time.

76
00:05:11,220 --> 00:05:14,940
So just to do just played here.

77
00:05:15,330 --> 00:05:16,820
Just played here.

78
00:05:17,070 --> 00:05:23,820
And if you are again going to execute it, so if you are going to call ahead on your data now, we will

79
00:05:23,820 --> 00:05:28,700
again see or there are two new Collum as a driver and it's go down.

80
00:05:28,710 --> 00:05:30,990
It's called hour and arrival and it's called time.

81
00:05:30,990 --> 00:05:37,330
And it's minute has been added in your data as well as you don't have any column name as a travel escort

82
00:05:37,410 --> 00:05:37,700
time.

83
00:05:38,010 --> 00:05:39,960
So that's about power, the automation.

84
00:05:39,960 --> 00:05:45,390
That's the power of creating a function in Python as well as in any programming language.

85
00:05:45,570 --> 00:05:52,770
Now, what we have to do is we have to process, we have to pre-process our this duration column.

86
00:05:53,010 --> 00:05:56,700
So we will observe over here in this duration column.

87
00:05:57,000 --> 00:06:02,120
This duration is nothing but in form of hours and minutes.

88
00:06:02,280 --> 00:06:07,230
So whenever you are going to parse this feature to your machine learning model as yeah, this is my

89
00:06:07,230 --> 00:06:12,780
duration, but machine learning model isn't able to understand because you have deltour machine learning

90
00:06:12,780 --> 00:06:13,210
model.

91
00:06:13,390 --> 00:06:17,720
Yeah, I have that much power in duration and that minute in duration.

92
00:06:18,060 --> 00:06:26,100
So but I think that you will notice over here maybe at some places you don't have a minute, you don't

93
00:06:26,100 --> 00:06:27,030
have a half hour as well.

94
00:06:27,030 --> 00:06:29,830
It means you have to pre-process this column as well.

95
00:06:30,150 --> 00:06:37,950
So what I am going to do over here, so that is first, let's say I have to outrate on each and every

96
00:06:37,950 --> 00:06:42,330
duration to where I'm going to say wherever I don't have any Myners.

97
00:06:42,540 --> 00:06:48,300
So simply I'm going to append as you get older and wherever, I don't have any hour in such case I can

98
00:06:48,360 --> 00:06:49,650
spend zero hour with.

99
00:06:50,040 --> 00:06:53,070
So for this, what I'm going to do, I'm just going to exit.

100
00:06:53,070 --> 00:07:02,100
As for I, in range of whatever will be the length of my duration column so that if I'm going to say

101
00:07:02,310 --> 00:07:08,080
this duration is nothing but I'm going to create a list of my duration.

102
00:07:08,370 --> 00:07:13,040
So just create a list of the so very first to have to add to the data.

103
00:07:13,230 --> 00:07:15,510
And here you have to access your duration.

104
00:07:15,720 --> 00:07:24,990
And after that you have to I try to zero to whatever will be the length of your duration list, whatever

105
00:07:25,170 --> 00:07:26,670
the length of my duration list.

106
00:07:27,280 --> 00:07:30,710
And once you have all these things that I'm going to say.

107
00:07:30,930 --> 00:07:32,370
So let me think.

108
00:07:32,380 --> 00:07:34,410
Let me show you a thing.

109
00:07:34,800 --> 00:07:40,710
Let's say I'm going to copy this and let's say I'm going to convert it into some string.

110
00:07:40,710 --> 00:07:48,330
And if I'm going to call a spade over there, let's say I have to split it on the basis of the space

111
00:07:48,360 --> 00:07:48,940
operator.

112
00:07:49,500 --> 00:07:52,420
So remove the extra spaces.

113
00:07:52,590 --> 00:08:00,780
So if I'm going to execute it now, we will see it will exactly the list where I have two hours and

114
00:08:00,990 --> 00:08:01,720
this minute.

115
00:08:01,890 --> 00:08:04,290
So my logic will be simple as that.

116
00:08:04,500 --> 00:08:12,430
Wherever the count, whatever it's count, whatever the count of the list will be two will be two.

117
00:08:12,570 --> 00:08:18,000
I don't have to do any manipulation in my data, but wherever it's length will be one.

118
00:08:18,180 --> 00:08:21,540
I have to do some modifications, some manipulations in my data.

119
00:08:21,720 --> 00:08:34,710
So here I'm going say is length of is length of my duration of I got a split because you have to split

120
00:08:34,710 --> 00:08:38,220
it and you have to split it on the basis of the space tiebreaker.

121
00:08:38,430 --> 00:08:43,110
And wherever this land is equally close to two, do so.

122
00:08:43,110 --> 00:08:45,870
In such case, what you have to do, you have to simply skip it.

123
00:08:45,870 --> 00:08:52,830
So to skip you guys can use this pass given order and here you have s block.

124
00:08:52,830 --> 00:08:57,170
And in this block, what you have to do, you have to wait for the condition.

125
00:08:57,480 --> 00:09:00,990
So wherever you have edge, it means that you have over.

126
00:09:01,200 --> 00:09:08,190
So wherever you have this hour in duration of I, it means in each and every duration, if you have

127
00:09:08,190 --> 00:09:10,910
this hour, if you have this edge.

128
00:09:11,220 --> 00:09:16,800
So in such case you have to append minute, you have to open the zero minute in your data.

129
00:09:16,920 --> 00:09:27,080
So here I'm going to say duration of by plus zero and you have to update this duration as well.

130
00:09:27,330 --> 00:09:36,030
So I'm going to say just based on the what if you don't have any edge, it means maybe you have a minute

131
00:09:36,030 --> 00:09:36,690
in your data.

132
00:09:36,700 --> 00:09:44,470
So in such case, what you have to do in such case, basically you have to spend zero hour in your data.

133
00:09:44,490 --> 00:09:49,570
So in such case, you have to upend zero hour in your data for here.

134
00:09:49,590 --> 00:09:54,610
I'm basically going to append this zero edge in my data.

135
00:09:55,050 --> 00:09:59,790
So what I have to do now just executed all of this just gets executed and.

136
00:10:00,280 --> 00:10:07,480
I would have call had on Monday, the same night you see all the modifications, but you will think,

137
00:10:07,480 --> 00:10:15,160
no, it's not happened because you don't have assigned this duration list in your data frame because

138
00:10:15,370 --> 00:10:16,720
it is not updated.

139
00:10:16,750 --> 00:10:20,300
You are too slow to update your data.

140
00:10:20,410 --> 00:10:25,620
You have to insert this judicial list at this place.

141
00:10:25,750 --> 00:10:31,750
So for this, what I am going to do is I have to excuse my data frame, then I have to access the duration

142
00:10:31,900 --> 00:10:36,880
and here I have to assign this duration so you don't have to copy.

143
00:10:36,910 --> 00:10:43,690
So I have to just as you ordered and if again, I am going to execute it, you will see a manipulation

144
00:10:43,690 --> 00:10:45,360
has been happen in your data.

145
00:10:45,370 --> 00:10:48,560
You will see a zero minute has been added in your data.

146
00:10:48,820 --> 00:10:50,380
That's what I'm trying to show you.

147
00:10:50,410 --> 00:10:52,070
So that's all about the session.

148
00:10:52,090 --> 00:10:54,010
Hopefully we'll love the session very much.

149
00:10:54,340 --> 00:10:55,240
Thank you, guys.

150
00:10:55,270 --> 00:10:56,170
Have a nice day.

151
00:10:56,320 --> 00:10:57,160
Keep learning.

152
00:10:57,160 --> 00:10:58,030
Keep growing.

153
00:10:58,330 --> 00:10:59,230
Keep practicing.