Scribble what I used when using ipython in formatting pos data

Introduction

I had to format the messy data with bytes, so it's a memo at that time The scribble of what I did in R about January is here. import Import of familiar data analysis tool

However, since I often handle data that spans many files, glob is also imported !! pandas

• I'm sorry I won't write about numpy this time. It's just data formatting. It is better to use nampy array for names, header, etc.

Read file

data = pd.read_csv("file name.csv")

--pd.read_ file format () supports various file formats --header = -1 can eliminate header! --names = ['1', '2', '3'] etc.

datas = glob.glob('*')

If you have a large number of files, this will give you all the files in the directory.

In the case of pandas, the called file will be a DataFrame type instead of a numpy array.

Erase columns and rows

It is a stripping off of the disturbing part.

data.drop([1,2])
#Clear line
data.drop([1,2],axis=1)
#Erase columns

By doing this, you can erase the row and column.

Join columns and rows

pd.concat([data[1],data[0]])
#Join lines
pd.concat([data[1],data[0]],axis=1)
#Join columns

This is useful when you have a lot of files and data!

Only rows with the data you want

Make only the data you want!

data.query("1==2")

Now you can only have a column named 1 with a value of 2.

Stick with the same data

Indispensable when molding files to derive data relationships!

pd.merge(data1, data, on='Column name')

This will stick the data together with the same column values.

Finally

Patience is important for blunt data shaping! After that, you can do it at once by using for etc. It is easy to swap columns and rows well by creating an array of numpy and a file in the middle. Think about the data you want and work hard toward it. Thank you for reading my poor memo.