1
00:00:00,480 --> 00:00:04,450
What is the average or
mean attack for each type?

2
00:00:04,450 --> 00:00:08,021
And what is the count of Pokemon for
each type?

3
00:00:08,021 --> 00:00:10,263
To answer our last two questions,

4
00:00:10,263 --> 00:00:15,460
we will use the cat plot sub family of
estimation plots with bar and count plots.

5
00:00:17,520 --> 00:00:19,516
Let's start with the barplot.

6
00:00:19,516 --> 00:00:24,741
sns.catplot (kind='bar',

7
00:00:27,525 --> 00:00:34,848
data=pokemon, x='Type',

8
00:00:34,848 --> 00:00:39,558
and y='Attack').

9
00:00:39,558 --> 00:00:42,319
Oops, and
let's remember to fix the aspect.

10
00:00:45,360 --> 00:00:47,786
Aspect=2.

11
00:00:49,275 --> 00:00:50,962
Beautiful.

12
00:00:50,962 --> 00:00:56,000
Seaborn's bar plot shows the mean or
average values for each category of data.

13
00:00:57,010 --> 00:01:01,150
It is in the categorical estimate
sub family because it shows

14
00:01:01,150 --> 00:01:03,516
an estimate of central tendency.

15
00:01:03,516 --> 00:01:08,464
The black lines on each bar are error
bars that represent uncertainty,

16
00:01:08,464 --> 00:01:11,691
like if there are outliers in our
observations.

17
00:01:11,691 --> 00:01:18,020
As we can see here, for
water there is a mean of 70.

18
00:01:18,020 --> 00:01:23,067
For grass it looks to be just about 60,

19
00:01:23,067 --> 00:01:30,200
Electric 75, Rock maybe 70,
Ground 75 and so on.

20
00:01:30,200 --> 00:01:34,720
I'll record my observations and
I'll catch you after the break.

21
00:01:36,360 --> 00:01:42,130
So I've recorded my average attack by
type observations into a markdown table.

22
00:01:42,130 --> 00:01:44,505
This is what it looks like in raw form so

23
00:01:44,505 --> 00:01:48,430
that you can see how to
make a table in markdown.

24
00:01:48,430 --> 00:01:50,372
When I hit Shift Return,

25
00:01:50,372 --> 00:01:55,009
I get a neatly formatted table
in my Jupyter Notebook here.

26
00:01:55,009 --> 00:01:59,362
So average attack by type,
Water 85, Grass 60,

27
00:01:59,362 --> 00:02:03,730
Electric 75, Rock 70,
Ground 75, and so on.

28
00:02:05,670 --> 00:02:08,004
Let's answer our second question now.

29
00:02:08,004 --> 00:02:11,230
What is the count of Pokemon for
each type?

30
00:02:11,230 --> 00:02:14,002
If we want to count each type of Pokemon,

31
00:02:14,002 --> 00:02:17,434
we can change the kind
keyword argument to count.

32
00:02:17,434 --> 00:02:23,000
But unlike the bar plot, the count
plot only needs one axis of data.

33
00:02:23,000 --> 00:02:23,943
Let's look at that.

34
00:02:23,943 --> 00:02:26,706
sns.catplot.

35
00:02:28,594 --> 00:02:31,548
Kind=count.

36
00:02:33,180 --> 00:02:35,302
Data=pokemon.

37
00:02:38,870 --> 00:02:41,387
X='Type'.

38
00:02:41,387 --> 00:02:47,301
For the count plot, unlike the bar
plot we only need one axis of data.

39
00:02:47,301 --> 00:02:49,010
Let's run that.

40
00:02:49,010 --> 00:02:52,384
My cell was a markdown.

41
00:02:52,384 --> 00:02:54,392
So, let's change that to code.

42
00:02:56,273 --> 00:02:56,930
There we go.

43
00:02:58,950 --> 00:03:05,095
And whoops, I didn't adjust the aspect,
so let's do that there, too, aspect=2.

44
00:03:07,788 --> 00:03:10,147
Beautiful.

45
00:03:10,147 --> 00:03:13,973
So, if we want to count
each type of Pokemon,

46
00:03:13,973 --> 00:03:19,031
we only need one axis of data,
and I set the x-axis to type.

47
00:03:21,220 --> 00:03:24,937
The most common types of
Pokemon are normal and psychic.

48
00:03:24,937 --> 00:03:25,520
Nice.

49
00:03:26,720 --> 00:03:30,532
I'll make a table with a count
of all my observations and

50
00:03:30,532 --> 00:03:32,813
I'll catch you after the break.

51
00:03:32,813 --> 00:03:37,770
So I've made my table here
with counter Pokemon by type.

52
00:03:37,770 --> 00:03:44,790
I've got 4 Water, 4 Grass, 2 Electric,
2 Rock, 3 Ground, and so on.

53
00:03:44,790 --> 00:03:49,367
With the combination of vertical pipes and
dashes, I'm able to make tables and

54
00:03:49,367 --> 00:03:52,720
markdown, and you should be able to, too.

55
00:03:52,720 --> 00:03:58,152
It's a neat way to present your
data in these Jupyter notebooks,

56
00:03:58,152 --> 00:04:02,611
and it it's just helpful to
have the data at a glance.

57
00:04:02,611 --> 00:04:08,017
Now that we've answered our two questions,
What is the average mean attack for

58
00:04:08,017 --> 00:04:11,980
each type, and what is the count
of Pokemon for each type?

59
00:04:13,030 --> 00:04:18,263
Try making a bar plot that
finds the average defense for

60
00:04:18,263 --> 00:04:22,259
each type, and make a table in markdown.

61
00:04:22,259 --> 00:04:25,171
I'll leave this here so
you can see the format again,

62
00:04:25,171 --> 00:04:29,323
try making a table in markdown that
summarizes your findings from the plot.