tutorial68-linear_regression.py

# https://youtu.be/9CxJhQynU20

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns

df = pd.read_csv('data/cells.csv')
print(df)
plt.scatter(x="time", y="cells", data=df)


#Now let us define our x and y values for the model.
#x values will be time column, so we can define it by dropping cells
#x can be multiple independent variables which we will discuss in a different tutorial
#this is why it is better to drop the unwanted columns rather than picking the wanted column
#y will be cells column, dependent variable that we are trying to predict. 

x_df = df.drop('cells', axis='columns')
#Or you can pick columns manually. Remember double brackets.
#Single bracket returns as series whereas double returns pandas dataframe which is what the model expects.
#x_df=df[['time']]
print(x_df.dtypes)  #Prints as object when you drop cells or use double brackets [[]]
#Prints as float64 if you do only single brackets, which is not the right type for our model. 
y_df = df.cells

#SPlit data into training and test datasets so we can validate the model using test data

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(x_df, y_df, test_size=0.3, random_state=42)
#random_state can be any integer and it is used as a seed to randomly split dataset.
#By doing this we work with same test dataset evry time, if this is important.
#random_state=None splits dataset randomly every time



#For linear regression, Y=the value we want to predict
#X= all independent variables upon which Y depends. 
#3 steps for linear regression....
#Step 1: Create the instance of the model
#Step 2: .fit() to train the model or fit a linear model
#Step 3: .predict() to predict Y for given X values. 
# Step 4: Calculate the accuracy of the model. 

#TO create a model instance 
from sklearn import linear_model
model = linear_model.LinearRegression()  #Create an instance of the model.
model.fit(X_train, y_train)   #Train the model or fits a linear model

print(model.score(X_train, y_train))  #Prints the R^2 value, a measure of how well
#observed values are replicated by themodel. 


prediction_test = model.predict(X_test)    
print(y_test, prediction_test)
print("Mean sq. errror between y_test and predicted =", np.mean(prediction_test-y_test)**2)
# A MSE value of about 8 is not bad compared to average # cells about 250.


#Residual plot
plt.scatter(prediction_test, prediction_test-y_test)
plt.hlines(y=0, xmin=200, xmax=300)

#Plot would be useful for lot of data points