Aggiungere evoluzioni

Analisi predittiva intermedia in Python

Nele Verbiest

Senior Data Scientist @PythonPredictions

Perché le evoluzioni (1)

Perché le evoluzioni (2)

Perché le evoluzioni (3)

Perché le evoluzioni (4)

Perché le evoluzioni (5)

Aggiungere evoluzioni alla basetable (1)

# Reference dates
start_2017 = datetime.date(2017,1,1)
start_2016 = datetime.date(2016,1,1)
start_2015 = datetime.date(2015,1,1)
# Gifts last month and last year
gifts_2016 = gifts[
   (gifts["date"]<start_2017) 
    & (gifts["date"]>=start_2016)]

gifts_2015_and_2016 = gifts[
   (gifts["date"]<start_2017) 
    & (gifts["date"]>=start_2015)]

Aggiungere evoluzioni alla basetable (2)

# Number of gifts in these periods per donor
number_gifts_2016 = gifts_2016.groupby("id")["amount"].size().reset_index()
number_gifts_2016.columns = ["donor_ID", "number_gifts_2016"]

number_gifts_2015_and_2016 = 
    gifts_2015_and_2016 .groupby("id")["amount"].size().reset_index()
number_gifts_2015_and_2016.columns = ["donor_ID", "number_gifts_2015_and_2016"]

Aggiungere evoluzioni alla basetable (3)

# Add these numbers to the basetable
basetable = pd.merge(basetable, 
                     number_gifts_2016, 
                     on="donor_ID",
                     how = "left")
basetable = pd.merge(basetable, 
                     number_gifts_2015_and_2016, 
                     on="donor_ID",
                     how = "left")

# Calculate ratio of last month's and last year's average
basetable["ratio_2015_to_2015_and_2016"] = 
    basetable["number_gifts_2016"] / 
    basetable["number_gifts_2015_and_2016"]

Aggiungere evoluzioni alla basetable (4)

print(basetable.head())

donor_id number_gifts_2016 number_gifts_2015_and_2016 ratio_2015_to_2015_and_2016
1        Na                5                          Na
2        9                 12                         0.75
3        3                 6                          0.5

Esercitiamoci!

Analisi predittiva intermedia in Python