load.py

real = pd.read_csv("/content/drive/My Drive/News/True.csv")
fake = pd.read_csv("/content/drive/My Drive/News/Fake.csv")


#Shape of real news dataset
print("Real news: ",real.shape)
#Shape of fake news dataset
print("Fake News: ", fake.shape)

# Assigning a value of 0 for all real news data and placing it in the dataframe
class0 = []

# 21417 is the length of the real news dataframe
for i in range(21417):
  class0.append(0)

real.insert(4, "class", class0, True)

#Assigning a value of 1 for all fake news data and placing it in the dataframe
class1 = []

#23481 is the length of the fake news dataframe
for i in range(23481):
  class1.append(1)

fake.insert(4, "class", class1, True)

#Concactenating fake news and real news into 1 total dataset
total = pd.concat([real,fake])
total = total.sample(frac = 1)

total.head()