1
00:00:01,159 --> 00:00:03,044
What is the relationship
between attack and type?

2
00:00:03,044 --> 00:00:07,987
In our previous scatter plots and
histograms we tried to visualize

3
00:00:07,987 --> 00:00:12,414
the attack data according to
a categorical variable type.

4
00:00:12,414 --> 00:00:17,195
Seaborn has a family of categorical
plots that will help us make better

5
00:00:17,195 --> 00:00:18,477
visualizations.

6
00:00:18,477 --> 00:00:24,571
Common categorical plots include
strip plot, box plot and bar plot.

7
00:00:24,571 --> 00:00:29,417
Let's start with making your categorical

8
00:00:29,417 --> 00:00:33,596
scatterplot called the stripplot.

9
00:00:35,351 --> 00:00:38,626
Sns.stripplot.

10
00:00:38,626 --> 00:00:42,881
Data-pokemon, x=type, and y=attack.

11
00:00:46,464 --> 00:00:51,446
So now we have a stripplot with all
of our observations grouped by type.

12
00:00:51,446 --> 00:00:55,320
But notice how the x axis labels
are all smushed together.

13
00:00:58,009 --> 00:00:59,813
Seaborn has two families of functions.

14
00:00:59,813 --> 00:01:04,186
Stripplot is an axis level function.

15
00:01:04,186 --> 00:01:09,254
In order to fix the x axis labels,

16
00:01:09,254 --> 00:01:13,779
we'll have to use a different

17
00:01:13,779 --> 00:01:18,847
family of function called figure

18
00:01:18,847 --> 00:01:23,915
level functions sns.catplot,

19
00:01:23,915 --> 00:01:30,774
kind='strip', Data=pokemon.

20
00:01:30,774 --> 00:01:31,881
X=Type.

21
00:01:31,881 --> 00:01:33,410
And y=Attack.

22
00:01:33,410 --> 00:01:38,495
When I run, Shift+Enter and
create this plot,

23
00:01:38,495 --> 00:01:46,943
I'll get a similar looking plot where my
x axis labels are all smushed together.

24
00:01:46,943 --> 00:01:49,753
However, when I use catplot,

25
00:01:49,753 --> 00:01:54,943
I have access to an additional
parameter called aspect,

26
00:01:54,943 --> 00:01:59,086
which will let me adjust
the x axis labels.

27
00:01:59,086 --> 00:01:59,745
Aspect=2.

28
00:02:01,934 --> 00:02:07,282
There we go, now I have a categorical
plot of the kind strip, the stripplot.

29
00:02:07,282 --> 00:02:12,314
And I can see all of
the distributions of my Pokemon's

30
00:02:12,314 --> 00:02:16,651
attack statistics grouped by their types.

31
00:02:16,651 --> 00:02:18,400
In this visualization,

32
00:02:18,400 --> 00:02:23,744
we can observe that the dragon type
Pokemon have the highest attack points.

33
00:02:23,744 --> 00:02:27,767
While the normal type Pokemon
have the largest spread of data.

34
00:02:27,767 --> 00:02:31,526
So let's record these observations
in a new mark down cell.

35
00:02:34,283 --> 00:02:38,559
The Dragon types have the highest attack.

36
00:02:43,587 --> 00:02:51,345
While the normal types,
Have the largest spread of data.

37
00:02:53,523 --> 00:02:56,013
These are just some observations
that I'm recording.

38
00:02:56,013 --> 00:02:57,643
Feel free to add more of your own.

39
00:02:57,643 --> 00:03:03,340
Now, try practicing making your stripplot
using the Pokemon's defense stat.