[PYTHON] [Super Basics] About jupyter Notebook

What is this?

I'm neither a software engineer nor a data scientist, but as I extract data from MySQL, BigQuery, etc. in my daily work, I've become interested in the method of statistically expressing and visualizing them. I think jupyter notebook is suitable for realizing them, and recently wrote Python in jupyter notebook. This article is a compilation of memorandums when I use jupyter, and the scope is from reading CSV data using pandas to checking basic statistics and simple data visualization methods.

This is an article written by a super amateur.

What is Jupyter Notebook?

Refer to official document

Code memo

What is Pandas? ... A library that provides functions to support data analysis. In particular, it provides data structures and operations for manipulating mathematical tables and time series data. What is numpy? ・・・ Library for numerical analysis What is pyplot? ・・・ Visualization library

1. Import data and check data

`test.ipynb`


# 1.Import the libraries needed for data analysis

import pandas as pd 
import numpy as np
from matplotlib import pyplot as plt
%matplotlib inline

#2.Read data * It is assumed that there is a column. If not header=Specify None as an option
# head()、tail()By using the function, you can check the contents of the data at the beginning or the end. I use it when there are many lines
data = pd.read_csv("hogehoge.csv") 

#3.Check the matrix of data
data.shape # -> (Rows, columns) will be returned.

#4.Check the basic statistics (basic characteristics of the data. Mean, standard deviation, maximum, minimum, etc.) and data type.
#Also, mean()By using functions such as, it is also possible to return only the average.
data.describe()
data.info()


#To specify a column: "Hogehoge" in the where clause in SQL=Image to specify "hoge"
data["hoge"]

2. Data visualization

`test.ipynb`



#Draw a line graph.
#The size can be specified as an option. figsize=(Horizontal size,Vertical size)Specify with
#In addition, the title can be specified as an option. title= "hoge"
data["hoge"].plot()

#Specifies the name of the x-axis and y-axis
label = date.plot(figzize=(15,5),title="test")
label.set_xlabel("hogehoge")
label.set_ylabel("hogehoge")

#variable.plot.hist()でヒストグラムを、variable.boxplot(by=x axis)Box plot is also possible with.