Sigmoid回归实现分类-带gui显示，下载就能使用

import numpy as np import matplotlib.pyplot as plt import pandas import wx class logistic: def __sigmoid(x): return 1.0/(1+np.exp(-x)) def Loaddata(name ="new.csv"): with open(name,'r',newline='') as f: lines=np.array(pandas.read_csv(f)).astype(np.float) #矩阵化 label=lines[:,4] #切片 label=np.expand_dims(label,axis=-1) data=lines[:,0:2]/129 N,W=np.shape(data) data=np.hstack((np.ones([N,1]),data)) return label,data def __WEight_updata(data,label,w,alpha=0.01): z=np.dot(data,w) h=logistic.__sigmoid(z) Error=label-h w=w+alpha*np.dot(data.T,Error) return w def train_LR(data,label,n_epoch=2,alpha=0.01): N,D=np.shape(data) w=np.ones([D,1]) #生成w for i in range (n_epoch): w=logistic.__WEight_updata(data,label,w,alpha) #print(w) error_rate=logistic.__test_accuracy(data,label,w) shuchu2.SetValue(str(error_rate)) return w def __test_accuracy(data,label,w): N,D=np.shape(data) z=np.dot(data,w) h=logistic.__sigmoid(z) lab_det=(h>0.5).astype(np.float) error_rate=np.sum(np.abs(label-lab_det))/N #print(N) return error_rate def huitu(data,label,w,*z): # print(w) z=np.array(z)/129 dic_colors={0:(.8,0,0),1:(0,.8,0),2:(0,0,0)} for i in range (2): index=np.where(label==i)[0] sub_data=data[index] plt.scatter(sub_data[:,1],sub_data[:,2],color=dic_colors[i]) plt.scatter(z[:,0],z[:,1],color=dic_colors[2]) min_x=np.min(data[:,1]) max_x=np.max(data[:,1]) w=w[:,0] x=np.arange(min_x,max_x,0.01) y=(-x*w[1]-w[0])/w[2] plt.plot(x,y) plt.show() def huitu2(data,label,w): # print(w) dic_colors={0:(.8,0,0),1:(0,.8,0),2:(0,0,0)} for i in range (2): index=np.where(label==i)[0] sub_data=data[index] plt.scatter(sub_data[:,1],sub_data[:,2],color=dic_colors[i]) min_x=np.min(data[:,1]) max_x=np.max(data[:,1]) w=w[:,0] x=np.arange(min_x,max_x,0.01) y=(-x*w[1]-w[0])/w[2] plt.plot(x,y) plt.show() def ceshi1(even): label,data=logistic.Loaddata() weights=logistic.train_LR(data,label,500,1.2) #print(weights) sg=float(shuru1.GetValue()) tz=float(shuru2.GetValue()) ceshi=(sg,tz) logistic.huitu(data,label,weights,ceshi) def ceshi2(even): label,data=logistic.Loaddata() weights=logistic.train_LR(data,label,500,1.2) logistic.huitu2(data,label,weights) app=wx.App() win=wx.Frame(None,title="logistic分类器",size=(600,600)) panel=wx.Panel(win) win.Show() huitu1Button=wx.Button(panel,label="测试",pos=(400,180)) huituButton=wx.Button(panel,label="查看样本分布",pos=(400,140)) shuchu1=wx.StaticText(panel,pos=(10,50),label=("错误率")) shuchu2=wx.TextCtrl(panel,pos=(50,50),style=wx.TE_READONLY,size=(240,30)) shuru1=wx.TextCtrl(panel,pos=(50,100),style=wx.TE_RICH) shuru2=wx.TextCtrl(panel,pos=(50,130),style=wx.TE_RICH) sgshuru=wx.StaticText(panel,pos=(10,100),label="身高") tzshuru=wx.StaticText(panel,pos=(10,130),label="体重") tishi0=wx.StaticText(panel,pos=(10,230),label="提示:") tishi1=wx.StaticText(panel,pos=(10,260),label="1,红色类,代表女生") tishi1.SetForegroundColour("red") tishi1.SetFont(wx.Font(13,wx.SWISS,wx.NORMAL,wx.NORMAL)) tishi2=wx.StaticText(panel,pos=(10,290),label="2,绿色类，代表男生") tishi2.SetForegroundColour("green") tishi2.SetFont(wx.Font(13,wx.SWISS,wx.NORMAL,wx.NORMAL)) tishi3=wx.StaticText(panel,pos=(10,320),label="3,黑色类，代表输入个人数据") tishi3.SetFont(wx.Font(13,wx.SWISS,wx.NORMAL,wx.NORMAL)) tishi3.SetForegroundColour("black") huitu1Button.Bind(wx.EVT_BUTTON,ceshi1) huituButton.Bind(wx.EVT_BUTTON,ceshi2) app.MainLoop()