How does Matplotlib generate multiple pictures in turn & only save these pictures without displaying them in the compiler

Catalog

1. Introduction to import library

2. Image generation code introduction

---

1. Introduction to import library

import os
import inspect
import pandas as pd
import numpy as np
from datetime import datetime

#  stay jupyter notebook Display image in 
import matplotlib.pyplot as plt
%matplotlib inline

#  Show Chinese in the picture 、 Minus sign, etc 
plt.rcParams['font.sans-serif']=['SimHei']
plt.rcParams['axes.unicode_minus']=False

import matplotlib
#  stay Python The generated image is not displayed in 
matplotlib.use('Agg')

Before the first part 5 Libraries involve data processing , The second part jupyter notebook Specific settings , The third part is about Chinese drawing names 、 Display of legend and minus sign , The fourth part focuses on the generated multiple pictures 、 Not displayed in the compiler .

2. Image generation code introduction

#  Variable initialization settings 
emer_sort = pd.DataFrame(columns = ['STARTTIME','WEEK','TIME','MINUTE','BELONGEDLINE','STATIONNAMES'])

#  Classify according to a column of data 
emer_stagroup = emer.groupby('STATIONNAMES')

#  Count effective data days k, Prepare for drawing multiple subgraphs 
k = 0
for sta_name, sta_group in emer_stagroup:
    #  according to STARTTIME Column sorting 
    emer1 = sta_group.sort_values(by = 'STARTTIME')
    emer_timegroup = emer1.groupby('TIME')
    for time_name, time_group in emer_timegroup:
        #  Select a date with sufficient sample size TIME - 5min Spaced samples ,192 yes 16 Hours 
        if time_group.shape[0] < 192:
            continue
        else:
            #  Add all sample points for a day 
            emer_sort = emer_sort.append(time_group)
            
            k = k + 1
            #  longitudinal 、 Horizontal axis data 
            time_group_flow = time_group['FLOW']
            time_group_minute = time_group['MINUTE'].to_list()

            fig, ax = plt.subplots(figsize = (10, 7))
            #  Axis scale font settings 
            plt.tick_params(labelsize=13)
            ax.plot(time_group_flow.values)
            #  Axis title settings 
            ax.set_xlabel('Date-time', fontsize=14)
            ax.set_ylabel('Flow', fontsize=14)
            #  Title Setting 
            ax.set_title(time_group['STATIONNAMES'].values[0], fontsize=14)
            #  Abscissa scale setting  -  How to display the date 
            dates = pd.to_datetime(time_group['MINUTE'])
            dates = dates.apply(lambda d: d.strftime('%H:%M'))
            ax.set_xticks(np.arange(len(dates))[12::24])
            #  Abscissa scale from 12 The number begins 、 every other 24 The number shows , tilt 45°
            _ = ax.set_xticklabels(dates[12::24], rotation=45)
            #  Storage 
            plt.savefig('./fig/fig_lots/flow_%s.png'%k, dpi=200)
            
#  Reset dataframe The line of label , and drop The previous label 
emer_sort_set = emer_sort.reset_index(drop = True)

If the program reports an error “Out of bounds Datetime: Out of bounds nanosecond timestamp: ...”, And point to this line of code  dates = pd.to_datetime(time_group['MINUTE']), The current solution is to Comment out this line and the following line .（ The question is Another article Some discussion , But there is no conclusion >_<）

Welcome to exchange and discuss ~