Code debug help # your code here allowed_factors = ['Age', 'Weight', 'Height', 'Neck', 'Chest', 'Abdomen', 'Hip',

No answer yet for this question. Ask a Tutor

Question:

Code debug help

# your code here
allowed_factors = ['Age', 'Weight', 'Height', 'Neck', 'Chest', 'Abdomen', 'Hip', 'Thigh', 'Knee', 'Ankle', 'Biceps', 'Forearm','Wrist']
# Initialize the model with no predictors
best = ['', 0]

class Model_Null:
def __init__(self, response):
self.endog_names = response
self.exog_names = ["Intercept"]

class Model_Null_Regression:
def __init__(self, response):
self.model = Model_Null(response)
self.rsq_adjusted = 0

def generate_formula(last_model, next_feature):
formula = last_model.endog_names + "~"
for i in range(len(last_model.exog_names)-1):
formula = formula + last_model.exog_names[i+1] + "+"
formula = formula + next_feature
return formula

def forward_stepwise_selection(last_model, allowed_factors, df):
current_best = (None, None, 0)
remaining_features = [feature for feature in allowed_factors if feature not in last_model.model.exog_names[1:]]

for feature in remaining_features:
model = smf.ols(generate_formula(last_model.model, feature), data=df).fit(disp=0)
if model.rsquared_adj > current_best[2]:
current_best = (feature, model, model.rsquared_adj)
return current_best

train_bmi1 = forward_stepwise_selection(Model_Null_Regression('Density'), allowed_factors, train_fat)[1]

train_bmi2 = forward_stepwise_selection(train_bmi1, allowed_factors, train_fat)[1]
train_bmi3 = forward_stepwise_selection(train_bmi2, allowed_factors, train_fat)[1]
train_bmi4 = forward_stepwise_selection(train_bmi3, allowed_factors, train_fat)[1]
train_bmi5 = forward_stepwise_selection(train_bmi4, allowed_factors, train_fat)[1]

3i. Plot [5 pts]
Plot your resulting adjusted ????2
vs number of predictors (k=1,2,3,4,5) and overlay the adjusted ????2
for the test data. Call the list of the five adjusted r-squared values from the five train_bmi# models as adjr2_train and the one from the test data as adjr2_test.# plot resulting adjusted rsquared vs number of predictors (k=1,2,3,4,5)

# Adjusted R^2 Formula = 1- [(1-R^2)*(n-1)/(n-k-1)]; R^2 sample R-squared, N: Total Sample Size, p: # of Independent variables

# When deep=True, data is copied but actual Python objects will not be copied recursively, only the reference to the object.
#This is in contrast to copy.deepcopy in the Standard Library, which recursively copies object data.
test_df = df.copy(deep=True)
#test_df.columns.drop("mpg")

def calculate_adjusted_rsq(test_df, linear_model):

response = linear_model.model.endog_names
features = linear_model.model.exog_names[1:]

y_test = test_df[response] # response = y

yhat_test = linear_model.predict(test_df[features]) # y_hat = predicted value of y

residual_squared_test = (y_test - yhat_test)**2

RSS_test = residual_squared_test.sum() # Residual Sum of Squares RSS

mean_test = y_test.mean() # y_bar aka mean

deviance_squared_test = (y_test-mean_test)**2

TSS_test = deviance_squared_test.sum() # Total Sum of Squares TSS

# R^2 = 1-RSS/TSS
Rsq_test = 1- (RSS_test/TSS_test)

n= len(y_test) # total sample size
k= len(features)

adjustedRsq_test = 1-[(1-Rsq_test)*(n-1)/(n-k-1)]

return adjustedRsq_test

k = [1, 2, 3, 4, 5]
k1 = forward_stepwise_selection(Model_Null_Regression("Density"), allowed_factors, train_fat)
k2 = forward_stepwise_selection(train_bmi1, allowed_factors, train_fat)
k3 = forward_stepwise_selection(train_bmi2, allowed_factors, train_fat)
k4 = forward_stepwise_selection(train_bmi3, allowed_factors, train_fat)
k5 = forward_stepwise_selection(train_bmi4, allowed_factors, train_fat)

adjr2_train = [k1[2], k2[2], k3[2], k4[2], k5[2]]

#set up an initial value for adjr2_test
adjr2_test = []

adjr2_test.append(calculate_adjusted_rsq(test_df, train_bmi1))
adjr2_test.append(calculate_adjusted_rsq(test_df, train_bmi2))
adjr2_test.append(calculate_adjusted_rsq(test_df, train_bmi3))
adjr2_test.append(calculate_adjusted_rsq(test_df, train_bmi4))
adjr2_test.append(calculate_adjusted_rsq(test_df, train_bmi5))

print(adjr2_train)
print(adjr2_test)

I am trying to code above code, but encountered an error, please help me debug.