I will introduce the continuation of the last time and the points I want to improve. Well, the figure was made like this. (Last time, I couldn't use Japanese for the axis name because I didn't understand the setting of "matplotlib". Since I corrected it this time, the axis name in the figure can be displayed in Japanese.)
Now let's play with this data a little more. This is a continuation of the previous code.
nikkei.corr()
Then you can get a table of correlations.
ax = sns.heatmap(nikkei.corr(), annot=True)
Furthermore, you can create a heat map. Like this...
There are few, but this time the code part is over.
Aside from the technical side, ** why I chose the heat map ... ** is a very big reflection. As you can see, the correlation coefficient is almost 1. What on earth are you doing? You should do this with a different subject. The subject of the Nikkei 225 is just a taste. I will post that I wish I could do it from the next time onward, including improvements. It is undecided whether to do it, but ...
--Display the Nikkei average on a candle chart --Try plotting NY Dow --Let's do regression analysis --Try adding another feature
And so on. The content is thin at the beginner level, but I want to do my best to continue.
Recommended Posts