|
通过wordcloud库、matplotlib库、networkx 库实现四大名著人物词云、出场次数统计图、人物社交关系网络图
你需要掌握的知识： 1.三大库 2.正则表达式 3.tkinter库 4.jieba库 5.python基本知识
效果如下：
 源代码：
import re
import jieba
import wordcloud
import matplotlib.pyplot as plt
from matplotlib.backends.backend_tkagg import FigureCanvasTkAgg,NavigationToolbar2Tk
from imageio import imread
import networkx as nx
import matplotlib
import tkinter as tk
from tkinter import ttk
from tkinter import *
class Graph_make:
def __init__(self):
self.root=Tk()
self.root.title('tk')
self.root.geometry('1000x800')
self.root.iconbitmap('0.gif')
self.Image_info=LabelFrame(self.root)
self.Image_info.place(x=100,y=320)
self.Image_info1=LabelFrame(self.root)
self.Image_info1.place(x=550,y=320)
self.first=True
self.Deal()
self.root.mainloop()
def Deal(self):
self.List()
self.Menu()
number=tk.StringVar()
self.Listchoose=ttk.Combobox(self.root,width=12,textvariable=number,state='readonly')
self.Listchoose['values']=book
self.Listchoose.bind("",self.show_msg)
self.Listchoose.pack()
self.Listchoose.set('四大名著')
self.root.mainloop()
def show_msg(self,*arg):
self.Displayall(self.Listchoose.get()+'.txt')
def Menu(self):
self.v=StringVar(self.root)
self.v.set('四大名著')
om=OptionMenu(self.root,self.v,*book)
om.bind('',self.submenu)
om.focus_set()
om.pack(side='left')
def submenu(self,event):
if self.v.get()=='四大名著':
return
self.Deal_data(self.v.get()+'.txt')
self.draw(self.v.get()+'.txt')
return
def draw(self,filepath):
print('键盘操作')
self.root2=Tk()
self.root2.geometry('1000x800')
self.menubar=tk.Menu(self.root2)
self.filemenu=tk.Menu(self.menubar,tearoff=0)
self.menubar.add_cascade(label='exit',menu=self.filemenu)
self.filemenu.add_command(label='exit',command=self.Exit)
self.root2.config(menu=self.menubar)
v=tk.IntVar()
self.labelframe=LabelFrame(self.root2,text=self.v.get())
self.labelframe.pack(side='left')
self.labelframe1=LabelFrame(self.root2,text=self.v.get(),padx=20,pady=10)
self.labelframe1.pack(fill="both", expand=True)
r1=tk.Radiobutton(self.labelframe,text='《'+filepath[:-4]+'》'+'词云',variable=v,value=1,command=self.wordcloud)
r1.pack(anchor="w")
r2=tk.Radiobutton(self.labelframe,text='《'+filepath[:-4]+'》主要人物社交关系网络图',variable=v,value=2,command=self.networks)
r2.pack(anchor="w")
r3=tk.Radiobutton(self.labelframe,text=filepath[:-4]+'主要人物统计图',variable=v,value=3,command=self.chart)
r3.pack(anchor="w")
self.root2.mainloop()
def Exit(self):
self.root2.destroy()
def wordcloud(self):
self.labelframe1.destroy()
self.labelframe1=LabelFrame(self.root2,text='《'+self.v.get()+'》'+'词云',padx=20,pady=10)
self.labelframe1.pack(fill="both", expand=True)
self.Wordcloud(self.v.get()+'.txt',self.labelframe1)
self.root2.update()
def networks(self):
self.labelframe1.destroy()
self.labelframe1=LabelFrame(self.root2,text='《'+self.v.get()+'》主要人物社交关系网络图',padx=20,pady=10)
self.labelframe1.pack(fill="both", expand=True)
self.Networks(self.v.get()+'.txt',self.labelframe1)
self.root2.update()
def chart(self):
self.labelframe1.destroy()
self.labelframe1=LabelFrame(self.root2,text=self.v.get()+'主要人物统计图',padx=20,pady=10)
self.labelframe1.pack(fill="both", expand=True)
self.Chart(self.v.get()+'.txt',self.labelframe1)
self.root2.update()
def Deal_data(self,filepath):
global relation,cnt
relation.clear()
cnt.clear()
plt.rcParams['font.sans-serif']=['SimHei']
with open(filepath,encoding='utf-8') as file:
text=file.read()
file.close()
words=jieba.lcut(text.strip())
stopwords=[line.strip() for line in open('停用表.txt','r',encoding='utf-8').readlines()]
counts={}
for word in words:
if len(word)==1:
continue
elif word not in stopwords:
counts[word]=counts.get(word,0)+1
items=list(counts.items())
items.sort(key=lambda x:x[1],reverse=True)
for i in range(24):
word,count=items[i]
cnt[word]=count
name=[]
for i in range(24):
name.append(items[i][0])
relation={}
lst_para=text.split('\n')
for txt in lst_para:
for name1 in name:
if name1 in txt:
for name2 in name:
if name2 in txt and name1!=name2 and (name2,name1) not in relation:
relation[(name1,name2)]=relation.get((name1,name2),0)+1
maxRela=max([v for k,v in relation.items()])
relation={k:v/maxRela for k,v in relation.items()}
def Displayall(self,filepath):
print('鼠标操作')
if self.first:
self.button=Button(self.root,width=8,height=2,text="Back",bg="orange",font=("楷", 12),command=self.Back)
self.button.place(x=0,y=550)
self.first=False
self.Deal_data(filepath)
self.Chart(filepath,self.root)
self.Wordcloud(filepath,self.Image_info)
self.Networks(filepath,self.Image_info1)
def Back(self):
self.Image_info.place_forget()
self.Image_info1.place_forget()
self.root1.destroy()
self.button.destroy()
self.first=True
self.root.update()
def Convert(self,para1,para2):
self.canvas1=FigureCanvasTkAgg(para1,para2)
self.canvas1.draw()
cv1=self.canvas1.get_tk_widget()
cv1.pack(side=tk.TOP, fill=tk.X, expand=1)
toolbar=NavigationToolbar2Tk(self.canvas1,para2)
toolbar.update()
self.canvas1._tkcanvas.pack(side=tk.TOP, fill=tk.BOTH,expand=1)
def Wordcloud(self,filepath,pos):
if pos==self.Image_info:
self.Image_info.destroy()
self.Image_info=LabelFrame(self.root)
self.Image_info.place(x=100,y=320)
pos=self.Image_info
self.root.update()
Figure=plt.figure(num='词云')
plt.title('《'+filepath[:-4]+'》'+'词云',color='white',fontweight='heavy',bbox=dict(facecolor='blue',edgecolor='r',alpha=0.65))
bg_pic=imread('词云背景图.jpg')
wcloud=wordcloud.WordCloud(mask=bg_pic,font_path=r'C:\Windows\Fonts\simhei.ttf',background_color='white',max_words=100,margin=1).fit_words(cnt)
plt.imshow(wcloud)
plt.axis('off')
self.Convert(Figure,pos)
def Networks(self,filepath,pos):
if pos==self.Image_info1:
self.Image_info1.destroy()
self.Image_info1=LabelFrame(self.root)
self.Image_info1.place(x=550,y=320)
pos=self.Image_info1
self.root.update()
self.figure=plt.figure()
G=nx.Graph()
plt.title(filepath[:-4],fontsize='large',color='white',fontweight='heavy',bbox=dict(facecolor='y',edgecolor='r',alpha=0.65))
for k,v in relation.items():
G.add_edge(k[0],k[1],weight=v)
elarge=[(u,v) for (u,v,d) in G.edges(data=True) if d['weight']>0.6]
emidle=[(u,v) for (u,v,d) in G.edges(data=True) if (d['weight']>0.3) & (d['weight'] |