rt-ai-4

'''
#Задание 1
import numpy as np
from numpy import *
from numpy.random import *
from scipy import linalg
import matplotlib.pyplot as plt

#Экстерполяция полинома 1-ой степени
x=np.array([0,1,2,3,4,5])
y=np.array([-3,-1.2,0.8,2.4,3.2,5])

A=np.vstack([x,np.ones(len(x))]).T
print(A)

m,c=np.linalg.lstsq(A,y,rcond=None)[0]
print(m,c)

plt.plot(x,y,'o',label='Исходные данные',markersize=10)
plt.plot(x,m*x+c,'r',label='Линейная экстраполяция')
plt.legend()
plt.show()

#Экстерполяция полинома 2-ой степени
delta=1.0
x=linspace(-8,8,18)
y=x**2+delta*(rand(18)-0.5)
x+=delta*(rand(18)-0.5)

x.tofile('x_data.txt','\n')
y.tofile('y_data.txt','\n')

x=fromfile('x_data.txt',float,sep='\n')
y=fromfile('y_data.txt',float,sep='\n')

print(x)
print(y)

m=vstack((x**2,x,ones(18))).T

s=np.linalg.lstsq(m,y,rcond=None)[0]

x_prec=linspace(-8,8,164)

plt.plot(x,y,'D')
plt.plot(x_prec,s[0]*x_prec**2+s[1]*x_prec+s[2],'-',lw=2)
plt.grid()
plt.savefig('парабола.png')

#Экстерполяция полинома 3-ей степени
m=vstack((x**3,x**2,x,ones(18))).T

s=np.linalg.lstsq(m,y,rcond=None)[0]

x_prec=linspace(-8,8,164)

plt.plot(x,y,'D')
plt.plot(x_prec,s[0]*x_prec**3+s[1]*x_prec**2+s[2]*x_prec+s[3],'-',lw=3)
plt.grid()
plt.savefig('парабола.png')

#Задание 2
import matplotlib.pyplot as plt
from scipy import linalg
import scipy as sp
from scipy.optimize import curve_fit
import numpy as np
from numpy import *
from numpy.random import *

#Функция 1
beta = (0.5, 1)
def f(x, b0, b1):
    return b0 + b1 * x

xdata = np.linspace(0, 8, 80)

y = f(xdata, *beta)

ydata = y + 0.05 * np.random.randn(len(xdata))
beta_opt, beta_cov = sp.optimize.curve_fit(f, xdata, ydata)
print (beta_opt)

lin_dev = sum(beta_cov[0])
print(lin_dev)

residuals = ydata - f(xdata,*beta_opt)
fres = sum(residuals**2)

print(fres)

fig, ax = plt.subplots()
ax.scatter(xdata, ydata)
ax.plot(xdata, y, 'r', lw=2)
ax.plot(xdata, f(xdata, *beta_opt), 'b', lw=2)
ax.set_xlim(0, 8)
ax.set_xlabel(r"$x$", fontsize=18)
ax.set_ylabel(r"$f(x, \beta)$", fontsize=18)
plt.show()

#Функция 2
beta = (0.5, 1.5, 1)
def f(x, b0, b1, b2):
    return b0+ b1 * x + b2*x*x

xdata = np.linspace(0, 8, 80)

y = f(xdata, *beta)

ydata = y + 0.05 * np.random.randn(len(xdata))
beta_opt, beta_cov = sp.optimize.curve_fit(f, xdata, ydata)
print(beta_opt)
lin_dev = sum(beta_cov[0])
print(lin_dev)

residuals = ydata - f(xdata, *beta_opt)
fres = sum(residuals**2)

fig, ax = plt.subplots()
ax.scatter(xdata, ydata)
ax.plot(xdata, y, 'r', lw=2)
ax.plot(xdata, f(xdata, *beta_opt), 'b', lw=2)
ax.set_xlim(0, 8)
ax.set_xlabel(r"$x$", fontsize=18)
ax.set_ylabel(r"$f(x, \beta)$", fontsize=18)
plt.show()

#Функция 3

beta = (2, 4)
def f(x, b0, b1):
    return b0 + b1 * np.log(x)

xdata = np.linspace(1, 8, 80)

y = f(xdata, *beta)

ydata = y + 0.05 * np.random.randn(len(xdata))
beta_opt, beta_cov = sp.optimize.curve_fit(f, xdata, ydata)
print(beta_opt)

lin_dev = sum(beta_cov[0])
print(lin_dev)


residuals = ydata - f(xdata, *beta_opt)
fres = sum(residuals**2)

print(fres)

fig, ax = plt.subplots()
ax.scatter(xdata, ydata)
ax.plot(xdata, y, 'r', lw=2)
ax.plot(xdata, f(xdata, *beta_opt), 'b', lw=2)
ax.set_xlim(0, 8)
ax.set_xlabel(r"$x$", fontsize=18)
ax.set_ylabel(r"$f(x, \beta)$", fontsize=18)
plt.show()

#Функция 4


beta = (2, 4)
def f(x, b0, b1):
    return b0 * x ** b1

xdata = np.linspace(1, 8, 80)

y = f(xdata, *beta)

ydata = y + 0.05 * np.random.randn(len(xdata))
beta_opt, beta_cov = sp.optimize.curve_fit(f, xdata, ydata)
print(beta_opt)

lin_dev = sum(beta_cov[0])
print(lin_dev)

residuals = ydata - f(xdata, *beta_opt)
fres = sum(residuals**2)

print(fres)

fig, ax = plt.subplots()
ax.scatter(xdata, ydata)
ax.plot(xdata, y, 'r', lw=2)
ax.plot(xdata, f(xdata, *beta_opt), 'b', lw=2)
ax.set_xlim(0, 8)
ax.set_xlabel(r"$x$", fontsize=18)
ax.set_ylabel(r"$f(x, \beta)$", fontsize=18)
plt.show()




#Задание 3
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from pandas import DataFrame,Series
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression

url="https://raw.githubusercontent.com/AnnaShestova/salary-years-simple-linear-regression/master/Salary_Data.csv"
df=pd.read_csv(url)

print(df.head(5))
print("--------------")

print(df.shape)
print("--------------")
print(df.describe())
print("--------------")

plt.scatter(df['YearsExperience'],df['Salary'],color='b',label="данные опыта работы и ЗП")
plt.xlabel("Опыт работы")
plt.ylabel("Заработная плата")
plt.show()

X=df.iloc[:,:-1].values
y=df.iloc[:,1].values
print(X)
print("--------------")
print(y)
print("--------------")

X_train,X_test,y_train,y_test=train_test_split(X,y,test_size=0.2,random_state=0)

regressor=LinearRegression()
regressor.fit(X_train,y_train)

print(regressor.intercept_)
print("--------------")
print(regressor.coef_)
print("--------------")

y_pred=regressor.predict(X_test)

datf=pd.DataFrame({'Actual':y_test,'Predicted':y_pred})
print(datf)
print("--------------")

plt.scatter(X_test,y_test,color='gray')
plt.plot(X_test,y_pred,color='red',linewidth=2)
plt.show()

'''

# Множественная линейная регрессия
import pandas as pd
from pandas import read_csv
import numpy as np
import seaborn as sn
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
from sklearn import metrics


url = r'https://raw.githubusercontent.com/aniruddhachoudhury/Red-Wine-Quality/master/winequality-red.csv'
df = read_csv(url)

df_pH = df['pH']

df_alc_qual = [df['alcohol'], df['quality']]

#print(df_pH.head(5), df_alc_qual.head(5), sep='\n')

new_df_pH = np.array(df_pH)
new_df_pH = new_df_pH.transpose()

new_df_alc_qual = np.array(df_alc_qual)
new_df_alc_qual = new_df_alc_qual.transpose()


df1 = pd.DataFrame(new_df_pH)
df2 = pd.DataFrame(new_df_alc_qual)

df1 = df1.rename(columns = {0: 'pH'}, inplace=False)
df2 = df2.rename(columns = {0: 'alcohol', 1: 'quality'}, inplace=False)


frames = [df1, df2]

dataset = pd.concat(frames, axis=1, join="inner")

print(dataset.head(5))

print(dataset.shape)

df_pH = dataset['pH']
df_alc_qual = dataset[['alcohol', 'quality']]

x_train, x_test, y_train, y_test = train_test_split(df_alc_qual, df_pH, test_size=0.2, random_state=42)

regressor = LinearRegression()
regressor.fit(x_train, y_train)

coeff_df = pd.DataFrame(regressor.coef_, df_alc_qual.columns, columns=['Coefficient'])

print(coeff_df)

y_pred = regressor.predict(x_test)

df_final = pd.DataFrame({'Actual': y_test, 'Predicted': y_pred})


print(df_final)

print('Mean Squared Error: ', metrics.mean_squared_error(y_test, y_pred))