Data visualization - White Snake 2: black snake robbery (1)

Catalog

Data loading and preprocessing

Premise ： Toolkit introduction

View

  Filter for null values

  Reset

Visual analysis

1、 Grade distribution

2、 Daily Comments

3、 Comments per hour

---

Data loading and preprocessing

Premise ： Toolkit introduction

# Data processing 
import numpy as np
import pandas as pd

# visualization 
import matplotlib.pyplot as plt
import seaborn as sns

from pyecharts.charts import Map
from pyecharts import options as opts
from pyecharts.globals import ThemeType, SymbolType, ChartType

Read the form   White Snake ( Including provinces ) data .xlsx

df = pd.read_excel(r"C:\Users\1\Desktop\ Data analysis \ White Snake ( Including provinces ) data .xlsx")

---

View

# It is used to view statistical indicators for numerical types
df.describle()

# View all field details
df.info()

# View table structure It can be used
df.head()

  adopt any() Know the null value in the table

Code runs ： 

array = np.array([True,False,False])
array.any()

  Operation result diagram ：

 

adopt mean() Know the null value in the table
df.isnull().mean()

Code runs ： 

df.isnull().mean()

Result chart ： 

---

  Filter for null values

It can be done by how and axis To formulate the rules of null value filtering

Filter null values with dropna(how='any',axis=0)
When how='any' If there is an empty row in the table, filter it out
When how='all' If this row in the table is empty, filter it out
axis = 0 Filter line
axis = 1 Filter column

Rule setting ： Delete a row of data with null values

Code ：df = df.dropna(how='any',axis=0)

df = df.dropna(how='any',axis=0)

  Reset

Because the index is filtering null values , There is a change , So reset
df.index = np.arange(df.shape[0])

---

Visual analysis

1、 Grade distribution

The goal is ： take   White Snake ( Including provinces ) data .xlsx The grade distribution in the table file is displayed graphically .

Global Sketchpad settings

sns.set （）

Parameter description ：
sort=True： Whether to sort ; Sort by default
ascending=False： Default descending order ;
normalize=False： Whether to standardize the calculation results , And display the results after standardization , The default is False.
bins=None： You can customize the grouping interval , Default whether ;
dropna=True： Delete missing values nan, Delete by default

bar Histogram

Code ： 

df_star = df[' score '].value_counts().sort_index(ascending=False)

sns.set()# Overall Sketchpad style    Background setting 
df_star.plot(kind='bar')

Result display ： 

Set the row coordinate cell to ‘ Numbers + branch  ’

Method 1：

def nonamefunction(x):
    return f'{x} branch ’

  Method 2：

lambda x : f'{x} branch '

Set the row coordinate cell to ‘ Numbers + branch  ’ Code ： 

df_star.index.map(lambda x : f'{x} branch ')

Result display ： 

  Configure the display of Chinese

plt.rcParams['font.sans-serif']='SimHei'
plt.rcParams['axes.unicode_minus']=False

Use histogram and pie chart to It means to view the data in the table , The code is as follows ：

plt.rcParams['font.sans-serif']='SimHei'
plt.rcParams['axes.unicode_minus']=False

figure = plt.figure(figsize=(12,5))#figsize=(12,5) wide 12 high 5

x = np.arange(df_star.size)
plt.bar(x,df_star.values)
# Custom scale 
_ = plt.xticks(x,df_star.index.map(lambda x : f'{x} branch '))

ax = figure.add_subplot(1,3,3)
_ = ax.pie(df_star.values,labels=df_star.index,autopct='%.1f%%')

Result display ： 

---

2、 Daily Comments

Show the daily comment time in the table ： 

df[' Comment on time '].min(),df[' Comment on time '].max()

Result chart ： 

  Set the time column to index , It is convenient to use the technology of time series to do data statistics

# Set the time column as the index , It is convenient to use the technology of time series to do data statistics 
df = df.set_index(' Comment on time ') 

 #D It means statistics by day 
comment_count = df.resample('D')[' Comment on '].coun

Recall the content of the last article here , Do a memory recovery

 plot usage , The code is as follows ：

text_x = np.linspace(0,2*np.pi)
text_y = np.sin(text_x)
plt.plot(text_x,text_y)

  Result display ：

 

Then we start Show the data volume of daily comments in August in a graph , The code is as follows ：

plt.rcParams['font.sans-serif']='SimHei'
plt.rcParams['axes.unicode_minus']=False
plt.figure(figsize=(12,5))
plt.plot(comment_count.index.day.tolist(),comment_count.values,color = 'green',marker = 'o')

for x,y in zip(comment_count.index.day.tolist(),comment_count.values):
    plt.text(x,y*1.08,str(y))

plt.title(' Daily comments in August ',fontsize=16,color='green')

plt.fill_between(comment_count.index.day.tolist(),comment_count.values,color='green')
_ = plt.xticks(comment_count.index.day.tolist())

  Result display ：

3、 Comments per hour

df.resample('H')[' Comment on '].count() #H  Hours 

df.reset_index(inplace = True)# Remember to run once   Do not repeat this sentence 
df[' Hours '] = df[' Comment on time '].dt.hour
comment_hours = df.groupby(' Hours ')[' Comment on '].count()

figure = plt.figure(figsize=(12,5))
ax1 = figure.add_subplot(1,1,1)
ax1.bar(comment_hours.index,comment_hours.values)
_ = ax1.set_xticks(comment_hours.index)
_ = ax1.set_xticklabels(comment_hours.index.map(lambda x:f'{x} when '))# Make one map mapping 

for x,y in zip(comment_hours.index,comment_hours.values):
    ax1.text(x-0.4,y*1.05,str(y))

plt.title(' Comments per hour ')
plt.xlabel(' Hours ')
plt.ylabel(' Comment frequency ')

  Follow up on the next ： Data visualization -《 White Snake 2： The green snake robbed 》（2）