def test_lm_fit(self):
my_lm = lm("y~.", data=df1)
my_lm.fit()
self.assertEqual(my_lm.intercept, 1.0)
self.assertEqual(my_lm.coefs[1], 3.5)
my_lm = lm("y~.-1", data=df2)
my_lm.fit()
self.assertEqual(my_lm.coefs[0], 1.5)
def test_lm_predict_c(self):
"""
This is further extension to the project, making sure the not implemented error is raised
"""
predict_df = pd.DataFrame({"x": [0, 1]})
my_lm = lm("y~.", data=df1)
my_lm.fit()
self.assertRaises(NotImplementedError,
lambda: my_lm.predict(predict_df, interval='c'))
def test_lm_summary(self):
"""
testing if the summary returned has the correct formula used for the model
"""
formula = "y~."
my_lm = lm(formula, data=df1)
my_lm.fit()
summary = my_lm.summary(print_summary=False)
lm_summary_formula = summary.splitlines()[2].split(',')[0].split(
'=')[1].strip()
self.assertEqual(formula, lm_summary_formula)
def test_lm_predict(self):
"""
Testing the predicted values for the linear model, to make sure the closed form solution works as expected
"""
predict_df = pd.DataFrame({"x": [0, 1]})
expected_y = [1, 4.5]
my_lm = lm("y~.", data=df1)
my_lm.fit()
y_hat = my_lm.predict(predict_df, interval=None)
self.assertListEqual(y_hat.tolist(), expected_y)
def test_lm_plot_all(self):
my_lm = lm("y~.", data=df1)
my_lm.fit()
self.assertRaises(NotImplementedError, lambda: my_lm.plot(which=None))
def test_lm_plot_3(self):
""" Testing to see if the plot was created and no error happened while creating the plot"""
my_lm = lm("y~.", data=df1)
my_lm.fit()
my_lm.plot(which=3)
import pandas as pd
from src.utility.FormulaParser import FormulaParser
from src.models.lm import lm
from src.models.knnRegressor import knnRegressor
from src.models.knnClassifier import knnClassifier
from src.models.BaseModel import BaseModel
if __name__ == "__main__":
model_formula = FormulaParser('dist~speed', ["speed", "dist"])
df = pd.DataFrame({"x": [0, 0, 1, 1], "y": [0, 2, 4, 5]})
print('Data to fit the model: \n', df)
my_lm = lm("y~.", data=df)
my_lm.fit()
my_lm.summary()
#
df1 = pd.DataFrame({"x": [0, 0, 1, 1], "y": [0, 2, 4, 5]})
predict_df = pd.DataFrame({"x": [0, 1]})
my_lm = lm("y~.", data=df1)
my_lm.fit()
my_lm.summary()
y_hat = my_lm.predict(predict_df)
print(y_hat)
my_knn = knnRegressor("y~.", data=df1)
my_knn.fit()
print(my_knn.predict(df1))
df2 = pd.DataFrame({"x": [0, 1, 2, 3], "y": [0, 0, 1, 1]})
