Codecademy export

cart.csv

script.py

import codecademylib
import pandas as pd
# -- inspecting the dataframes --
visits = pd.read_csv('visits.csv',
                     parse_dates=[1])
#print(visits) 
cart = pd.read_csv('cart.csv',
                   parse_dates=[1])
#print(cart)
checkout = pd.read_csv('checkout.csv',
                       parse_dates=[1])
#print(checkout)
purchase = pd.read_csv('purchase.csv',
                       parse_dates=[1])
#print(purchase)
# ----------------------------------------------------
def mergeCal(df,df2):
  merged = pd.merge(df,df2,how='left') #left merge on df and df2
  info_col = merged.columns[-1] # the _time column
  null_rows = merged[merged[info_col].isnull()] # null rows 
  percentage = float(len(null_rows))/len(merged)*100 # dividing number of null rows by number of rows in merged df 
  return merged,percentage,len(null_rows)

visitsCart,visitsCart_per,visitsCart_null = mergeCal(visits,cart)
print(visitsCart.shape)
#print(visitsCart_per) 82.6% of users didn't place a t-shirt in their cart

cartCheckout,cartCheckout_per,cartCheckout_null = mergeCal(cart,checkout)
#print(cartCheckout_per) 25.3% of users did not proceed to checkout

all_data = visits.merge(cart,how='left')\
           .merge(checkout,how='left')\
           .merge(purchase,how='left')

checkoutPurchase,checkoutPurchase_per,checkoutPurchase_null = mergeCal(checkout,purchase)
#print(checkoutPurchase_per) 16.8% of users did not purchase a t-shirt

# the weakest step of the funnel is the purchase step. Cool T-Shirts could look at how well-recieved their prices are, and survey for best-selling t-shirt qualities

all_data['time_to_purchase'] = \
    all_data.purchase_time - \
    all_data.visit_time

print(all_data.time_to_purchase.mean())

visits.csv