40 lines
1.4 KiB
Python
40 lines
1.4 KiB
Python
import pandas as pd
|
|
import numpy as np
|
|
|
|
data = pd.read_csv('./googleplaystore.csv')
|
|
|
|
data.dropna(subset=['Rating', 'Type','Content Rating','Current Ver','Android Ver'], inplace=True)
|
|
data.reset_index(drop=True, inplace=True)
|
|
data = data.drop(columns=["Size", "Android Ver", "Current Ver", "Last Updated"])
|
|
|
|
# normalizing text
|
|
to_lowercase = ['App', 'Category', 'Type', 'Content Rating', 'Genres']
|
|
for column in to_lowercase:
|
|
data[column] = data[column].apply(str.lower)
|
|
|
|
data["Installs"] = data["Installs"].replace({'\+': ''}, regex=True)
|
|
data["Installs"] = data["Installs"].replace({',': ''}, regex=True)
|
|
|
|
# normalizing numbers
|
|
data["Reviews"] = pd.to_numeric(data["Reviews"], errors='coerce')
|
|
max_value = data["Reviews"].max()
|
|
min_value = data["Reviews"].min()
|
|
data["Reviews"] = (data["Reviews"] - min_value) / (max_value - min_value)
|
|
|
|
data["Installs"] = pd.to_numeric(data["Installs"], errors='coerce')
|
|
max_value = data["Installs"].max()
|
|
min_value = data["Installs"].min()
|
|
data["Installs"] = (data["Installs"] - min_value) / (max_value - min_value)
|
|
|
|
#print(data)
|
|
|
|
|
|
# splitting into sets
|
|
np.random.seed(123)
|
|
train, validate, test = np.split(data.sample(frac=1, random_state=42), [int(.6*len(data)), int(.8*len(data))])
|
|
print(f"Data shape: {data.shape}\nTrain shape: {train.shape}\nTest shape: {test.shape}\nValidation shape:{validate.shape}")
|
|
|
|
train.to_csv('appstrain.csv')
|
|
test.to_csv('appstest.csv')
|
|
validate.to_csv('appsvalidate.csv')
|