Script 773: Pacing New Campaign Calculations

Purpose

This Python script solves the problem of calculating various metrics and values related to campaign pacing and budget allocation.
To Elaborate

The problem being solved is to calculate and analyze the performance and pacing of different campaigns based on their start and end dates, goals, and actual performance data. The script calculates metrics such as total pub cost, total impressions, total clicks, total views, auto pacing cycle start and end dates, auto pacing cycle days, auto pacing cycle pub cost, auto pacing cycle impressions, auto pacing cycle clicks, auto pacing cycle views, total target spend per impressions/views, total days, total days elapsed, total daily target, total expected to date, total pacing, delivery status, auto pacing cycle target remaining, pacing cycle daily allocation, and recommended daily budget.
Walking Through the Code

The script starts by importing the necessary libraries and defining the necessary constants.
It then creates two dataframes, df1 and df2, by selecting specific columns from the input dataframe inputDf.
The script converts the date columns in df1 and df2 to datetime format.
It then iterates through each row in df1 and performs the following steps:
- Filters df2 based on the campaign, account, and date range.
- Calculates the sum of pub cost, clicks, impressions, and views for the filtered data and assigns them to new columns in df1.
The script drops duplicate rows based on the ‘Campaign’ column in df1 and assigns the deduplicated dataframe to df1_deduplicated.
It then calculates the auto pacing cycle start date, auto pacing cycle end date, total days, auto pacing cycle pub cost, auto pacing cycle impressions, auto pacing cycle clicks, auto pacing cycle views, today’s date, auto pacing cycle days, auto pacing cycle days elapsed, auto pacing cycle days remaining, auto pacing cycle daily target spend per impressions/views, auto pacing cycle expected to date, auto pacing cycle pacing, auto pacing cycle threshold, total target spend per impressions/views, total days elapsed, total daily target, total expected to date, total pacing, delivery status, auto pacing cycle target remaining, pacing cycle daily allocation, and recommended daily budget for each row in df1_deduplicated.
Finally, the script prints the updated dataframe df1_deduplicated and assigns it to outputDf.
Vitals

Script ID : 773
Client ID / Customer ID: 1306927189 / 60270139
Action Type: Bulk Upload (Preview)
Item Changed: Campaign
Output Columns: Account, Campaign, Auto. Pacing Cycle Clicks, Auto. Pacing Cycle Days, Auto. Pacing Cycle Days Elapsed, Auto. Pacing Cycle Days Remaining, Auto. Pacing Cycle End Date, Auto. Pacing Cycle Expected to Date, Auto. Pacing Cycle Impr., Auto. Pacing Cycle Pacing, Auto. Pacing Cycle Pub. Cost, Auto. Pacing Cycle Start Date, Auto. Pacing Cycle Threshold, Auto. Pacing Cycle Views, Delivery Status, Recommended Daily Budget, Todays Date, Total Clicks, Total Daily Target, Total Days, Total Days Elapsed, Total Expected to Date, Total Impressions, Total Pacing, Total Pub Cost, Total Target (Spend/Impr./Views), Total Views
Linked Datasource: M1 Report
Reference Datasource: None
Owner: ascott@marinsoftware.com (ascott@marinsoftware.com)
Created by ascott@marinsoftware.com on 2024-03-06 22:19
Last Updated by ascott@marinsoftware.com on 2024-03-19 18:50
> See it in Action

Python Code

## Infinite Digital Pacing Script
## name: Script - New Campaign Calculations
## description:
##  
## 
## author: Jesus A. Garza
## created: 2024-02-06
## 

today = datetime.datetime.now().date()

# primary data source and columns
inputDf = dataSourceDict["1"]
RPT_COL_CAMPAIGN = 'Campaign'
RPT_COL_DATE = 'Date'
RPT_COL_ACCOUNT = 'Account'
RPT_COL_BRAND = 'Brand'
RPT_COL_GOAL = 'Goal'
RPT_COL_PACING__START_DATE = 'Pacing - Start Date'
RPT_COL_PACING__END_DATE = 'Pacing - End Date'
RPT_COL_TARGET_IMPR_PER_SPENDVIEWS = 'Target (Impr/Spend/Views)'
RPT_COL_CPM_RESTRAINT = 'CPM Restraint'
RPT_COL_CPC_RESTRAINT = 'CPC Restraint'
RPT_COL_CAMPAIGN_TYPE = 'Campaign Type'
RPT_COL_CAMPAIGN_STATUS = 'Campaign Status'
RPT_COL_SBA_TRAFFIC = 'SBA Traffic'
RPT_COL_TOTAL_PUB_COST = 'Total Pub Cost'
RPT_COL_TOTAL_IMPRESSIONS = 'Total Impressions'
RPT_COL_TOTAL_CLICKS = 'Total Clicks'
RPT_COL_TOTAL_VIEWS = 'Total Views'
RPT_COL_PACING_CYCLE_PUB_COST = 'Pacing Cycle Pub Cost'
RPT_COL_PACING_CYCLE_IMPRESSIONS = 'Pacing Cycle Impressions'
RPT_COL_PACING_CYCLE_CLICKS = 'Pacing Cycle Month Clicks'
RPT_COL_PACING_CYCLE_VIEWS = 'Pacing Cycle Views'
RPT_COL_AUTO_PACING_CYCLE_START_DATE = 'Auto. Pacing Cycle Start Date'
RPT_COL_AUTO_PACING_CYCLE_END_DATE = 'Auto. Pacing Cycle End Date'
RPT_COL_TODAYS_DATE = 'Todays Date'
RPT_COL_AUTO_PACING_CYCLE_DAYS = 'Auto. Pacing Cycle Days'
RPT_COL_AUTO_PACING_CYCLE_DAYS_ELAPSED = 'Auto. Pacing Cycle Days Elapsed'
RPT_COL_AUTO_PACING_CYCLE_DAYS_REMAINING = 'Auto. Pacing Cycle Days Remaining'
RPT_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS = 'Auto. Pacing Daily Target (Spend/Impressions/Views)'
RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE = 'Auto. Pacing Cycle Expected to Date'
RPT_COL_AUTO_PACING_CYCLE_PACING = 'Auto. Pacing Cycle Pacing'
RPT_COL_AUTO_PACING_CYCLE_THRESHOLD = 'Auto. Pacing Cycle Threshold'
RPT_COL_AUTO_PACING_CYCLE_CLICKS = 'Auto. Pacing Cycle Clicks'
RPT_COL_AUTO_PACING_CYCLE_IMPR = 'Auto. Pacing Cycle Impr.'
RPT_COL_AUTO_PACING_CYCLE_PUB_COST = 'Auto. Pacing Cycle Pub. Cost'
RPT_COL_AUTO_PACING_CYCLE_VIEWS = 'Auto. Pacing Cycle Views'
RPT_COL_TOTAL_TARGET_SPEND_PER_IMPRVIEWS = 'Total Target (Spend/Impr./Views)'
RPT_COL_TOTAL_DAYS = 'Total Days'
RPT_COL_TOTAL_DAYS_ELAPSED = 'Total Days Elapsed'
RPT_COL_TOTAL_DAILY_TARGET = 'Total Daily Target'
RPT_COL_TOTAL_EXPECTED_TO_DATE = 'Total Expected to Date'
RPT_COL_TOTAL_PACING = 'Total Pacing'
RPT_COL_DELIVERY_STATUS = 'Delivery Status'
RPT_COL_AUTO_PACING_CYCLE_TARGET_REMAINING_SPEND_PER_IMPRVIEWS = 'Auto. Pacing Cycle Target Remaining (Spend/Impr./Views)'
RPT_COL_AUTO_PACING_CYCLE_DAILY_ALLOCATION_SPEND_PER_IMPRVIEWS = 'Auto. Pacing Cycle Daily Allocation (Spend/Impr./Views)'
RPT_COL_RECOMMENDED_DAILY_BUDGET = 'Recommended Daily Budget'
RPT_COL_IMPR = 'Impr.'
RPT_COL_CLICKS = 'Clicks'
RPT_COL_PUB_COST = 'Pub. Cost $'
RPT_COL_VIDEO_VIEWS = 'Video Views'

# output columns and initial values
BULK_COL_ACCOUNT = 'Account'
BULK_COL_CAMPAIGN = 'Campaign'
BULK_COL_AUTO_PACING_CYCLE_CLICKS = 'Auto. Pacing Cycle Clicks'
BULK_COL_AUTO_PACING_CYCLE_DAYS = 'Auto. Pacing Cycle Days'
BULK_COL_AUTO_PACING_CYCLE_DAYS_ELAPSED = 'Auto. Pacing Cycle Days Elapsed'
BULK_COL_AUTO_PACING_CYCLE_DAYS_REMAINING = 'Auto. Pacing Cycle Days Remaining'
BULK_COL_AUTO_PACING_CYCLE_END_DATE = 'Auto. Pacing Cycle End Date'
BULK_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE = 'Auto. Pacing Cycle Expected to Date'
BULK_COL_AUTO_PACING_CYCLE_IMPR = 'Auto. Pacing Cycle Impr.'
BULK_COL_AUTO_PACING_CYCLE_PACING = 'Auto. Pacing Cycle Pacing'
BULK_COL_AUTO_PACING_CYCLE_PUB_COST = 'Auto. Pacing Cycle Pub. Cost'
BULK_COL_AUTO_PACING_CYCLE_START_DATE = 'Auto. Pacing Cycle Start Date'
BULK_COL_AUTO_PACING_CYCLE_TARGET_REMAINING_SPEND_PER_IMPRVIEWS = 'Auto. Pacing Cycle Target Remaining (Spend/Impr./Views)'
BULK_COL_AUTO_PACING_CYCLE_THRESHOLD = 'Auto. Pacing Cycle Threshold'
BULK_COL_AUTO_PACING_CYCLE_VIEWS = 'Auto. Pacing Cycle Views'
BULK_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS = 'Auto. Pacing Daily Target (Spend/Impressions/Views)'
BULK_COL_DELIVERY_STATUS = 'Delivery Status'
BULK_COL_AUTO_PACING_CYCLE_DAILY_ALLOCATION_SPEND_PER_IMPRVIEWS = 'Auto. Pacing Cycle Daily Allocation (Spend/Impr./Views)'
BULK_COL_RECOMMENDED_DAILY_BUDGET = 'Recommended Daily Budget'
BULK_COL_TODAYS_DATE = 'Todays Date'
BULK_COL_TOTAL_CLICKS = 'Total Clicks'
BULK_COL_TOTAL_DAILY_TARGET = 'Total Daily Target'
BULK_COL_TOTAL_DAYS = 'Total Days'
BULK_COL_TOTAL_DAYS_ELAPSED = 'Total Days Elapsed'
BULK_COL_TOTAL_EXPECTED_TO_DATE = 'Total Expected to Date'
BULK_COL_TOTAL_IMPRESSIONS = 'Total Impressions'
BULK_COL_TOTAL_PACING = 'Total Pacing'
BULK_COL_TOTAL_PUB_COST = 'Total Pub Cost'
BULK_COL_TOTAL_TARGET_SPEND_PER_IMPRVIEWS = 'Total Target (Spend/Impr./Views)'
BULK_COL_TOTAL_VIEWS = 'Total Views'
outputDf[BULK_COL_AUTO_PACING_CYCLE_CLICKS] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_CYCLE_DAYS] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_CYCLE_DAYS_ELAPSED] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_CYCLE_DAYS_REMAINING] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_CYCLE_END_DATE] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_CYCLE_IMPR] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_CYCLE_PACING] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_CYCLE_PUB_COST] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_CYCLE_START_DATE] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_CYCLE_TARGET_REMAINING_SPEND_PER_IMPRVIEWS] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_CYCLE_THRESHOLD] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_CYCLE_VIEWS] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS] = "<<YOUR VALUE>>"
outputDf[BULK_COL_DELIVERY_STATUS] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTO_PACING_CYCLE_DAILY_ALLOCATION_SPEND_PER_IMPRVIEWS] = "<<YOUR VALUE>>"
outputDf[BULK_COL_RECOMMENDED_DAILY_BUDGET] = "<<YOUR VALUE>>"
outputDf[BULK_COL_TODAYS_DATE] = "<<YOUR VALUE>>"
outputDf[BULK_COL_TOTAL_CLICKS] = "<<YOUR VALUE>>"
outputDf[BULK_COL_TOTAL_DAILY_TARGET] = "<<YOUR VALUE>>"
outputDf[BULK_COL_TOTAL_DAYS] = "<<YOUR VALUE>>"
outputDf[BULK_COL_TOTAL_DAYS_ELAPSED] = "<<YOUR VALUE>>"
outputDf[BULK_COL_TOTAL_EXPECTED_TO_DATE] = "<<YOUR VALUE>>"
outputDf[BULK_COL_TOTAL_IMPRESSIONS] = "<<YOUR VALUE>>"
outputDf[BULK_COL_TOTAL_PACING] = "<<YOUR VALUE>>"
outputDf[BULK_COL_TOTAL_PUB_COST] = "<<YOUR VALUE>>"
outputDf[BULK_COL_TOTAL_TARGET_SPEND_PER_IMPRVIEWS] = "<<YOUR VALUE>>"
outputDf[BULK_COL_TOTAL_VIEWS] = "<<YOUR VALUE>>"

print(inputDf.columns)

# Create DataFrame 1
data_df1 = inputDf[[RPT_COL_DATE, RPT_COL_CAMPAIGN, RPT_COL_ACCOUNT, RPT_COL_TARGET_IMPR_PER_SPENDVIEWS, RPT_COL_PACING__START_DATE, RPT_COL_PACING__END_DATE, RPT_COL_GOAL, RPT_COL_CLICKS, RPT_COL_IMPR, RPT_COL_CAMPAIGN_STATUS]].copy()

# Create DataFrame 2
data_df2 = inputDf[[RPT_COL_DATE, RPT_COL_ACCOUNT, RPT_COL_CAMPAIGN, RPT_COL_PUB_COST, RPT_COL_CLICKS, RPT_COL_IMPR, RPT_COL_VIDEO_VIEWS]]

# Create DataFrame 1 and DataFrame 2
df1 = pd.DataFrame(data_df1)
df2 = pd.DataFrame(data_df2)

# Convert 'Pacing - Start Date' and 'Pacing - End Date' columns to datetime
df1[RPT_COL_PACING__START_DATE] = pd.to_datetime(df1[RPT_COL_PACING__START_DATE])
df1[RPT_COL_PACING__END_DATE] = pd.to_datetime(df1[RPT_COL_PACING__END_DATE])
df2[RPT_COL_DATE] = pd.to_datetime(df2[RPT_COL_DATE])

# Populate the new columns in df1 with the sum of values from df2 for each campaign
for index, row in df1.iterrows():
    campaign = row[RPT_COL_CAMPAIGN]
    account = row[RPT_COL_ACCOUNT]
    start_date = row[RPT_COL_PACING__START_DATE]
    end_date = row[RPT_COL_PACING__END_DATE]
    
    # Filter df2 for the specific campaign and within the date range
    filtered_data = df2[(df2[RPT_COL_ACCOUNT] == account) & (df2[RPT_COL_CAMPAIGN] == campaign) & (df2[RPT_COL_DATE] >= start_date) & (df2[RPT_COL_DATE] <= end_date)]
    
    # Calculate the sum and assign them to the new columns in df1
    df1.loc[index, RPT_COL_TOTAL_PUB_COST] = filtered_data[RPT_COL_PUB_COST].sum().round(2)
    df1.loc[index, RPT_COL_TOTAL_CLICKS] = filtered_data[RPT_COL_CLICKS].sum().round(2)
    df1.loc[index, RPT_COL_TOTAL_IMPRESSIONS] = filtered_data[RPT_COL_IMPR].sum().round(2)
    df1.loc[index, RPT_COL_TOTAL_VIEWS] = filtered_data[RPT_COL_VIDEO_VIEWS].sum().round(2)

# Drop duplicates based on the 'Campaign' column
df1_deduplicated = df1.drop_duplicates(subset=[RPT_COL_CAMPAIGN]).copy()

# Display the updated DataFrame
print(df1_deduplicated)


##Todays date
df1_deduplicated[RPT_COL_TODAYS_DATE] = today

##Auto. Pacing Cycle Start Date 
##Formula: =IF(A2<>"",IF(AND(YEAR(E2)=YEAR(Y2),MONTH(Y2)=MONTH(E2)),E2,IF(Y2>F2,"",IF(AH2<=31,E2,EOMONTH(Y2,-1)+1))),"")
##Python
# Ensure the relevant columns are in datetime format
df1_deduplicated[RPT_COL_PACING__START_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_PACING__START_DATE], errors='coerce')
df1_deduplicated[RPT_COL_PACING__END_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_PACING__END_DATE], errors='coerce')
df1_deduplicated[RPT_COL_TODAYS_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_TODAYS_DATE], errors='coerce')

def calculate_auto_pacing_cycle_start_date(row):
    start_date = row['Pacing - Start Date'] 
    today_date = pd.to_datetime('today').normalize()  # Make today's date timezone naive
    end_date = row['Pacing - End Date']

    if pd.notnull(start_date) and pd.notnull(today_date):
        if start_date.year == today_date.year and start_date.month == today_date.month:
            result = start_date
        elif today_date > end_date:  
            return pd.NaT
        else:
            result = today_date.replace(day=1)  # Default to the first day of the current month
        
        return result
    else:
        return pd.NaT

# Apply the function to each row of the DataFrame
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_START_DATE] = df1_deduplicated.apply(calculate_auto_pacing_cycle_start_date, axis=1)


##Auto. Pacing Cycle End Date   
##Formula: =IF(A2<>"",IF(Y2>F2,"",IF(AND(YEAR(F2)=YEAR(Y2),MONTH(Y2)=MONTH(F2)),F2,IF(AH2<=31,E2,EOMONTH(Y2,0)))),"")
##Python
# Ensure the relevant columns are in datetime format
df1_deduplicated[RPT_COL_PACING__END_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_PACING__END_DATE], errors='coerce')
df1_deduplicated[RPT_COL_TODAYS_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_TODAYS_DATE], errors='coerce')
df1_deduplicated[RPT_COL_PACING__START_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_PACING__START_DATE], errors='coerce')

def calculate_auto_pacing_cycle_end_date(row):
    if pd.notnull(row['Campaign']):  # Check if campaign identifier is not empty
        end_date = row['Pacing - End Date']
        today_date = pd.to_datetime('today').normalize()  # Make today's date timezone naive
        
        if today_date > end_date:
            return ""
        elif end_date.year == today_date.year and end_date.month == today_date.month:
            return end_date
        else:
            # Use MonthEnd to get the end of the current month for today's date
            return pd.offsets.MonthEnd(0).rollforward(today_date)
    else:
        return pd.NaT  # Return Not a Time (NaT) for empty or invalid campaign identifiers

# Apply the function to each row of the DataFrame
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_END_DATE] = df1_deduplicated.apply(calculate_auto_pacing_cycle_end_date, axis=1)


##Total Days    
##Formula: =if(A2<>"",IF(F2-E2+1<1,"",F2-E2+1),"")
##Python 
# Convert date columns to datetime format
df1_deduplicated[RPT_COL_PACING__START_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_PACING__START_DATE], errors='coerce')
df1_deduplicated[RPT_COL_PACING__END_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_PACING__END_DATE], errors='coerce')

def calculate_total_days(row):
    if pd.notnull(row[RPT_COL_CAMPAIGN]):  
        # Calculate the difference in days, then add 1
        days_diff = (row[RPT_COL_PACING__END_DATE] - row[RPT_COL_PACING__START_DATE]).days + 1
        # Check if the calculated difference is less than 1
        if days_diff < 1:
            return 0  
        else:
            return days_diff
    else:
        return 0  

# Apply the function to each row in the DataFrame
df1_deduplicated[RPT_COL_TOTAL_DAYS] = df1_deduplicated.apply(calculate_total_days, axis=1)

##Auto. Pacing Cycle Pub. Cost 
##Formula: =if(A2<>"",sumIFs('Campaign Daily Data'!E:E,'Campaign Daily Data'!B:B,">="&W2,'Campaign Daily Data'!B:B,"<="&X2,'Campaign Daily Data'!A:A,A2),"")
##Python
# Convert date columns to datetime format if not already
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_START_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_START_DATE], errors='coerce')
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_END_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_END_DATE], errors='coerce')
df1_deduplicated[RPT_COL_TODAYS_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_TODAYS_DATE], errors='coerce')

# Calculate Auto. Pacing Cycle Pub Cost for each row in inputDf
def calculate_auto_pacing_cycle_pub_cost(row):
    if row[RPT_COL_CAMPAIGN] != "":
        start_date = row[RPT_COL_AUTO_PACING_CYCLE_START_DATE]
        end_date = row[RPT_COL_AUTO_PACING_CYCLE_END_DATE]
        campaign = row[RPT_COL_CAMPAIGN]

        # Filter 'campaign_daily_data' based on the criteria
        filtered_data = df2[
            (df2[RPT_COL_CAMPAIGN] == campaign) & 
            (df2[RPT_COL_DATE] >= start_date) & 
            (df2[RPT_COL_DATE] <= end_date)
        ]

        # Sum the publication costs for the filtered data
        total_pub_cost = filtered_data[RPT_COL_PUB_COST].sum()
        return total_pub_cost
    else:
        return 0 

# Apply the function to each row in 'inputDf' and create a new column for the auto pacing cycle publication cost
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_PUB_COST] = df1_deduplicated.apply(calculate_auto_pacing_cycle_pub_cost, axis=1)
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_PUB_COST] = df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_PUB_COST].round(2)


##Auto. Pacing Cycle Impr. 
##Formula: =if(A2<>"",sumIFs('Campaign Daily Data'!C:C,'Campaign Daily Data'!B:B,">="&W2,'Campaign Daily Data'!B:B,"<="&X2,'Campaign Daily Data'!A:A,A2),"") 
##Python 
# Convert date columns to datetime format if not already
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_START_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_START_DATE], errors='coerce')
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_END_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_END_DATE], errors='coerce')
df1_deduplicated[RPT_COL_TODAYS_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_TODAYS_DATE], errors='coerce')

# Calculate Auto. Pacing Cycle Impressions for each row in inputDf
def calculate_auto_pacing_cycle_impressions(row):
    if row[RPT_COL_CAMPAIGN] != "":
        start_date = row[RPT_COL_AUTO_PACING_CYCLE_START_DATE]
        end_date = row[RPT_COL_AUTO_PACING_CYCLE_END_DATE]
        campaign = row[RPT_COL_CAMPAIGN]

        # Filter 'campaign_daily_data' based on the criteria
        filtered_data = df2[
            (df2[RPT_COL_CAMPAIGN] == campaign) & 
            (df2[RPT_COL_DATE] >= start_date) & 
            (df2[RPT_COL_DATE] <= end_date)
        ]

        # Sum the impressions for the filtered data
        total_impressions = filtered_data[RPT_COL_IMPR].sum()
        return total_impressions
    else:
        return 0 

# Apply the function to each row in 'inputDf' and create a new column for the auto pacing cycle impressions
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_IMPR] = df1_deduplicated.apply(calculate_auto_pacing_cycle_impressions, axis=1)


##Auto. Pacing Cycle Clicks 
##Formula: =if(A2<>"",sumIFs('Campaign Daily Data'!D:D,'Campaign Daily Data'!B:B,">="&W2,'Campaign Daily Data'!B:B,"<="&X2,'Campaign Daily Data'!A:A,A2),"")
##Python 
# Define the column names based on your data source dictionary
# Convert date columns to datetime format if not already
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_START_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_START_DATE], errors='coerce')
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_END_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_END_DATE], errors='coerce')
df1_deduplicated[RPT_COL_TODAYS_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_TODAYS_DATE], errors='coerce')

# Calculate Auto. Pacing Cycle Clicks for each row in inputDf
def calculate_auto_pacing_cycle_clicks(row):
    if row[RPT_COL_CAMPAIGN] != "":
        start_date = row[RPT_COL_AUTO_PACING_CYCLE_START_DATE]
        end_date = row[RPT_COL_AUTO_PACING_CYCLE_END_DATE]
        campaign = row[RPT_COL_CAMPAIGN]

        # Filter 'campaign_daily_data' based on the criteria
        filtered_data = df2[
            (df2[RPT_COL_CAMPAIGN] == campaign) & 
            (df2[RPT_COL_DATE] >= start_date) & 
            (df2[RPT_COL_DATE] <= end_date)
        ]

        # Sum the clicks for the filtered data
        total_clicks = filtered_data[RPT_COL_CLICKS].sum()
        return total_clicks
    else:
        return 0  

# Apply the function to each row in 'inputDf' and create a new column for the auto pacing cycle clicks
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_CLICKS] = df1_deduplicated.apply(calculate_auto_pacing_cycle_clicks, axis=1)


##Auto. Pacing Cycle Views  
##Formula: =if(A2<>"",sumIFs('Campaign Daily Data'!F:F,'Campaign Daily Data'!B:B,">="&W2,'Campaign Daily Data'!B:B,"<="&X2,'Campaign Daily Data'!A:A,A2),"")
##Python 
# Convert date columns to datetime format if not already
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_START_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_START_DATE], errors='coerce')
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_END_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_END_DATE], errors='coerce')
df1_deduplicated[RPT_COL_TODAYS_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_TODAYS_DATE], errors='coerce')

# Calculate Auto. Pacing Cycle Views for each row in inputDf
def calculate_auto_pacing_cycle_views(row):
    if row[RPT_COL_CAMPAIGN] != "":
        start_date = row[RPT_COL_AUTO_PACING_CYCLE_START_DATE]
        end_date = row[RPT_COL_AUTO_PACING_CYCLE_END_DATE]
        campaign = row[RPT_COL_CAMPAIGN]

        # Filter 'campaign_daily_data' based on the criteria
        filtered_data = df2[
            (df2[RPT_COL_CAMPAIGN] == campaign) & 
            (df2[RPT_COL_DATE] >= start_date) & 
            (df2[RPT_COL_DATE] <= end_date)
        ]

        # Sum the views for the filtered data
        total_views = filtered_data[RPT_COL_VIDEO_VIEWS].sum()
        return total_views
    else:
        return 0  

# Apply the function to each row in 'inputDf' and create a new column for the auto pacing cycle views
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_VIEWS] = df1_deduplicated.apply(calculate_auto_pacing_cycle_views, axis=1)


##Todays Date   
##Formula: =if(A2<>"",today(),"")
##Python: 
#today = datetime.datetime.now(CLIENT_TIMEZONE).date()

##Auto. Pacing Cycle Days   
##Formula: =if(A2<>"",if(X2="","",if(AH2<=31,AH2,(X2-W2)+1)),"")
##Python:
# First, ensure that the dates are in the appropriate datetime format
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_START_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_START_DATE], errors='coerce')
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_END_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_END_DATE], errors='coerce')

def calculate_apc_days(row):
    if pd.notnull(row[RPT_COL_CAMPAIGN]) and pd.notnull(row[RPT_COL_AUTO_PACING_CYCLE_START_DATE]) and pd.notnull(row[RPT_COL_AUTO_PACING_CYCLE_END_DATE]):
        if row[RPT_COL_TOTAL_DAYS] <= 31:  
            return row[RPT_COL_TOTAL_DAYS]  
        else:
            # Ensure both dates are timezone-naive before subtracting
            start_date = row[RPT_COL_AUTO_PACING_CYCLE_START_DATE]
            end_date = row[RPT_COL_AUTO_PACING_CYCLE_END_DATE]
            return (end_date - start_date).days + 1
    else:
        return ""
        
# Apply the function row-wise
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_DAYS] = df1_deduplicated.apply(calculate_apc_days, axis=1)

##Auto. Pacing Cycle Days Elapsed
##Formula: =IF(A2<>"",IF(X2="","",IF(Y2-W2<=0,0,Y2-W2)),"")
##Definition: If campaign column is not blank and Auto. Pacing Cycle Start Date is not blank
##Python: 
# Convert the relevant columns to datetime format, assuming they're not already
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_START_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_START_DATE], errors='coerce')
df1_deduplicated[RPT_COL_TODAYS_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_TODAYS_DATE], errors='coerce')
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_END_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_END_DATE], errors='coerce')  # In case it's needed

def calculate_date_apc_days_elapsed(row):
    if pd.notnull(row[RPT_COL_CAMPAIGN]):  # Checks if the campaign column is not empty
        start_date = row[RPT_COL_AUTO_PACING_CYCLE_START_DATE]
        today_date = pd.to_datetime(today).normalize()  # Make today's date timezone naive
        
        if pd.isnull(start_date):  # Checks if the calculated start date is empty
            return ""
        # Calculate the difference between the dates in 'today_date' and 'start_date'
        diff = (today_date - start_date).days
        if diff <= 0:  # If the difference is less than or equal to 0
            return 0
        else:
            return diff
    else:
        return ""

# Apply the function to each row in the DataFrame
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_DAYS_ELAPSED] = df1_deduplicated.apply(calculate_date_apc_days_elapsed, axis=1)

    
##Auto. Pacing Cycle Days Remaining 
##Formula: =if(A2<>"",IF(Z2-AA2<0,0,Z2-AA2),"")
##Python: 
def calculate_apc_days_remaining(row):
    if pd.notnull(row[RPT_COL_CAMPAIGN]):
        # Convert the values to numeric types before subtraction
        days = pd.to_numeric(row[RPT_COL_AUTO_PACING_CYCLE_DAYS], errors='coerce')
        days_elapsed = pd.to_numeric(row[RPT_COL_AUTO_PACING_CYCLE_DAYS_ELAPSED], errors='coerce')
        
        # Check if the conversion resulted in NaN (not a number)
        if pd.isnull(days) or pd.isnull(days_elapsed):
            return ""  # Return an empty string or some other appropriate value for your context
        
        # Calculate the difference
        diff = days - days_elapsed
        return max(0, diff)
    else:
        return ""

# Apply the function to each row in the DataFrame
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_DAYS_REMAINING] = df1_deduplicated.apply(calculate_apc_days_remaining, axis=1)


##Auto. Pacing Daily Target (Spend, Impressions, Views) 
##Formula: =iferror(IF(D2="ms",G2/Z2,IF(D2="CPM",G2/Z2,IF(D2="CPV",G2/Z2,""))),"")
##Python
def calculate_auto_pacing_daily_target(row):
    try:
        # Since the calculation is the same for all specified campaign types, we directly perform the division
        if row[RPT_COL_GOAL] in ["MS", "CPM", "CPV"]:
            return row[RPT_COL_TARGET_IMPR_PER_SPENDVIEWS] / row[RPT_COL_AUTO_PACING_CYCLE_DAYS] if row[RPT_COL_AUTO_PACING_CYCLE_DAYS] != 0 else 0
        else:
            return 0  # Default to 0 for undefined campaign types or other errors
    except:
        return 0  # Safeguard against any other kind of error

# Apply the function to each row of the DataFrame
df1_deduplicated[RPT_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS] = df1_deduplicated.apply(calculate_auto_pacing_daily_target, axis=1)
df1_deduplicated[RPT_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS] = df1_deduplicated[RPT_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS].round(2)
df1_deduplicated[RPT_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS] = df1_deduplicated[RPT_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS].astype(str)


##Auto. Pacing Cycle Expected to Date   
##Formula: =if(A2<>"",AC2*AA2,"")
##Python:
def calculate_apc_etd(row):
    if pd.notnull(row[RPT_COL_CAMPAIGN]):
        # Convert both columns to numeric types, coercing errors to NaN
        daily_target = pd.to_numeric(row[RPT_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS], errors='coerce')
        days_elapsed = pd.to_numeric(row[RPT_COL_AUTO_PACING_CYCLE_DAYS_ELAPSED], errors='coerce')
        
        # Check if either value is NaN after conversion
        if pd.isnull(daily_target) or pd.isnull(days_elapsed):
            return 0  # Return 0 or some other appropriate default value
        
        # Perform the multiplication
        return daily_target * days_elapsed
    else:
        return 0  # Return 0 or some other appropriate default value if the campaign is null

# Apply the function to each row in the DataFrame
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE] = df1_deduplicated.apply(calculate_apc_etd, axis=1)
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE] = df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE].round(2)
#df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE] = df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE].astype(str)


##Auto. Pacing Cycle Pacing 
##Formula: =IFerror(IF(D2="ms",S2/AD2,IF(D2="CPM",T2/AD2,IF(D2="CPV",V2/AD2,0))),0)
##Python
def calculate_auto_pacing_cycle_pacing(row):
    try:
        # Check campaign type and perform corresponding calculation
        if row[RPT_COL_GOAL] == "MS":
            return row[RPT_COL_AUTO_PACING_CYCLE_PUB_COST] / row[RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE] if row[RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE] != 0 else 0
        elif row[RPT_COL_GOAL] == "CPM":
            return row[RPT_COL_AUTO_PACING_CYCLE_IMPR] / row[RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE] if row[RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE] != 0 else 0
        elif row[RPT_COL_GOAL] == "CPV":
            return row[RPT_COL_AUTO_PACING_CYCLE_VIEWS] / row[RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE] if row[RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE] != 0 else 0
        else:
            return 0  
    except:
        return 0  

# Calculate pacing as a percentage and convert to numeric type
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_PACING] = df1_deduplicated.apply(calculate_auto_pacing_cycle_pacing, axis=1)
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_PACING] = pd.to_numeric(df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_PACING], errors='coerce') * 100.0
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_PACING] = df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_PACING].round(2)
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_PACING] = df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_PACING].astype(str) + '%'


##Auto. Pacing Cycle Threshold  
##Formula: =if(A2<>"",IF(Y2>F2,"Campaign Ended",if(AE2<97%,"Under Pacing",if(AND(AE2<104%,AE2>103%),"Over Pacing",if(AE2>105%,"Severely Overpacing","On Pace")))),"")
##Python
# Convert the relevant columns to datetime format, assuming they're not already
df1_deduplicated[RPT_COL_PACING__END_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_PACING__END_DATE], errors='coerce')
df1_deduplicated[RPT_COL_TODAYS_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_TODAYS_DATE], errors='coerce')

def calculate_auto_pacing_cycle_threshold(row):
    today_date = row[RPT_COL_TODAYS_DATE]
    if pd.notnull(row[RPT_COL_CAMPAIGN]):  
        end_date = row[RPT_COL_PACING__END_DATE]

        if end_date is not pd.NaT and today_date > end_date:
            return "Campaign Ended"
        else:
            # Convert pacing to a float by removing the '%' sign and dividing by 100
            pacing_str = row[RPT_COL_AUTO_PACING_CYCLE_PACING].replace('%', '')  # Remove the '%' sign
            pacing = float(pacing_str) / 100.0  # Convert to float and divide by 100
            
            if pacing < 0.97: 
                return "Under Pacing"
            elif 0.97 <= pacing < 1.04:  
                return "On Pace"
            elif 1.04 <= pacing <= 1.05:  
                return "Over Pacing"
            elif pacing > 1.05:  
                return "Severely Overpacing"
            else:
                return "On Pace"
    else:
        return ""

# Apply the function to each row of the DataFrame
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_THRESHOLD] = df1_deduplicated.apply(calculate_auto_pacing_cycle_threshold, axis=1)


##Total Target (Spend/Impr./Views) 
##Formula: =if(A2<>"",IF(F2-E2+1<=31,G2,(G2/30)*(F2-E2)+1),"")
##Python
# Ensure date columns are in datetime format for calculation
df1_deduplicated[RPT_COL_PACING__START_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_PACING__START_DATE], errors='coerce')
df1_deduplicated[RPT_COL_PACING__END_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_PACING__END_DATE], errors='coerce')

def calculate_total_target(row):
    if row[RPT_COL_CAMPAIGN] != "":  
        # Calculate the duration from the start and end dates
        duration_days = (row[RPT_COL_PACING__END_DATE] - row[RPT_COL_PACING__START_DATE]).days + 1
        if duration_days <= 31:
            return row[RPT_COL_TARGET_IMPR_PER_SPENDVIEWS]
        else:
            return (row[RPT_COL_TARGET_IMPR_PER_SPENDVIEWS] / 30) * duration_days
    else:
        return ""

# Apply the function to each row of the DataFrame
df1_deduplicated[RPT_COL_TOTAL_TARGET_SPEND_PER_IMPRVIEWS] = df1_deduplicated.apply(calculate_total_target, axis=1)
df1_deduplicated[RPT_COL_TOTAL_TARGET_SPEND_PER_IMPRVIEWS] = df1_deduplicated[RPT_COL_TOTAL_TARGET_SPEND_PER_IMPRVIEWS].round(2)
df1_deduplicated[RPT_COL_TOTAL_TARGET_SPEND_PER_IMPRVIEWS] = df1_deduplicated[RPT_COL_TOTAL_TARGET_SPEND_PER_IMPRVIEWS].replace(np.nan, '', regex=True)
df1_deduplicated[RPT_COL_TOTAL_TARGET_SPEND_PER_IMPRVIEWS] = df1_deduplicated[RPT_COL_TOTAL_TARGET_SPEND_PER_IMPRVIEWS].astype(str)

##Total Days    
##Formula: =if(A2<>"",IF(F2-E2+1<1,"",F2-E2+1),"")
##Python 
# Convert date columns to datetime format
df1_deduplicated[RPT_COL_PACING__START_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_PACING__START_DATE], errors='coerce')
df1_deduplicated[RPT_COL_PACING__END_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_PACING__END_DATE], errors='coerce')

def calculate_total_days(row):
    if pd.notnull(row[RPT_COL_CAMPAIGN]):  
        # Calculate the difference in days, then add 1
        days_diff = (row[RPT_COL_PACING__END_DATE] - row[RPT_COL_PACING__START_DATE]).days + 1
        # Check if the calculated difference is less than 1
        if days_diff < 1:
            return ""
        else:
            return days_diff
    else:
        return ""

# Apply the function to each row in the DataFrame
df1_deduplicated[RPT_COL_TOTAL_DAYS] = df1_deduplicated.apply(calculate_total_days, axis=1)


##Total Days Elapsed  
##Formula: =IF(A2<>"",IF(AH2="","",IF(Y2-E2<=0,"",if(Y2-E2>=AH2,AH2,Y2-E2))),"")  
##Python
# Ensure the date columns are in datetime format
df1_deduplicated[RPT_COL_PACING__START_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_PACING__START_DATE], errors='coerce')
df1_deduplicated[RPT_COL_TODAYS_DATE] = pd.to_datetime(df1_deduplicated[RPT_COL_TODAYS_DATE], errors='coerce')

# Define the calculation as a function
def calculate_days_elapsed(row):
    if pd.notnull(row[RPT_COL_CAMPAIGN]):  # Checks if 'A' is not empty
        if pd.isnull(row[RPT_COL_TOTAL_DAYS]):  # Checks if 'AH' is empty
            return ""

        today_date = row[RPT_COL_TODAYS_DATE]
        start_date = row[RPT_COL_PACING__START_DATE]
        
        if start_date is not pd.NaT:
            start_date = pd.to_datetime(start_date.date())  # Convert to Timestamp
        today_date = row[RPT_COL_TODAYS_DATE]

        # Calculate the difference in days between 'Y' and 'E'
        days_diff = (today_date - start_date).days
        if days_diff <= 0:
            return ""
        # Convert RPT_COL_TOTAL_DAYS to an integer before comparison
        total_days = int(row[RPT_COL_TOTAL_DAYS]) if row[RPT_COL_TOTAL_DAYS] != "" else 0
        if days_diff >= total_days:
            return total_days
        else:
            return days_diff
    else:
        return ""

# Apply the function across each row
df1_deduplicated[RPT_COL_TOTAL_DAYS_ELAPSED] = df1_deduplicated.apply(calculate_days_elapsed, axis=1)


##Total Daily Target    
##Formula: =iferror(AG2/AH2,"")
##Python 
# Safe division function to avoid division by zero and handle errors
def safe_division(numerator, denominator):
    try:
        # Convert numerator to a numeric type, coercing errors to NaN
        numerator = pd.to_numeric(numerator, errors='coerce')
        # Check if the denominator is a numeric value and not zero
        if isinstance(denominator, (int, float)) and denominator != 0:
            return numerator / denominator
        else:
            return None  # or return 0 if that's more appropriate for your context
    except ZeroDivisionError:  # In case of division by zero
        return None

# Apply the safe division function to calculate the total daily target
df1_deduplicated[RPT_COL_TOTAL_DAILY_TARGET] = df1_deduplicated.apply(lambda row: safe_division(row[RPT_COL_TOTAL_TARGET_SPEND_PER_IMPRVIEWS], row[RPT_COL_TOTAL_DAYS]), axis=1)
df1_deduplicated[RPT_COL_TOTAL_DAILY_TARGET] = df1_deduplicated[RPT_COL_TOTAL_DAILY_TARGET].round(2)

##Total Expected to Date    
##Formula: =if(AI2<>"",AI2*AJ2,"")
##Python 
# Calculate the total expected to date using the conditions described
def calculate_expected_to_date(row):
    # Ensure that both values are numeric before multiplying
    total_days_elapsed = pd.to_numeric(row[RPT_COL_TOTAL_DAYS_ELAPSED], errors='coerce')
    total_daily_target = pd.to_numeric(row[RPT_COL_TOTAL_DAILY_TARGET], errors='coerce')
    
    # Check if either value is NaN after conversion (which could happen if coercion failed)
    if pd.isnull(total_days_elapsed) or pd.isnull(total_daily_target):
        return 0  # Return 0 or NaN if that's more appropriate for your context
    
    # Perform the multiplication
    return total_days_elapsed * total_daily_target

# Apply the function to each row in the DataFrame
df1_deduplicated[RPT_COL_TOTAL_EXPECTED_TO_DATE] = df1_deduplicated.apply(calculate_expected_to_date, axis=1)
df1_deduplicated[RPT_COL_TOTAL_EXPECTED_TO_DATE] = df1_deduplicated[RPT_COL_TOTAL_EXPECTED_TO_DATE].round(2)


##Total Pacing  
##Formula: =IFerror(IF(D2="ms",O2/AK2,IF(D2="CPM",P2/AK2,IF(D2="CPV",R2/AK2,""))),"")
##Python
# Function to calculate Total Pacing as a percentage
def calculate_total_pacing(row):
    try:
        # Goal type is 'ms'
        if row[RPT_COL_GOAL] == 'MS':
            return row[RPT_COL_TOTAL_PUB_COST] / row[RPT_COL_TOTAL_EXPECTED_TO_DATE] if row[RPT_COL_TOTAL_EXPECTED_TO_DATE] != 0 else ""
        # Goal type is 'CPM'
        elif row[RPT_COL_GOAL] == 'CPM':
            return row[RPT_COL_TOTAL_IMPRESSIONS] / row[RPT_COL_TOTAL_EXPECTED_TO_DATE] if row[RPT_COL_TOTAL_EXPECTED_TO_DATE] != 0 else ""
        # Goal type is 'CPV'
        elif row[RPT_COL_GOAL] == 'CPV':
            return row[RPT_COL_TOTAL_VIEWS] / row[RPT_COL_TOTAL_EXPECTED_TO_DATE] if row[RPT_COL_TOTAL_EXPECTED_TO_DATE] != 0 else ""
        else:
            return ""
    except:
        # Handles division by zero or any other calculation error
        return ""

# Calculate pacing as a percentage and convert to numeric type
df1_deduplicated[RPT_COL_TOTAL_PACING] = df1_deduplicated.apply(calculate_total_pacing, axis=1)
df1_deduplicated[RPT_COL_TOTAL_PACING] = pd.to_numeric(df1_deduplicated[RPT_COL_TOTAL_PACING], errors='coerce') * 100.0
df1_deduplicated[RPT_COL_TOTAL_PACING] = df1_deduplicated[RPT_COL_TOTAL_PACING].round(2)
df1_deduplicated[RPT_COL_TOTAL_PACING] = df1_deduplicated[RPT_COL_TOTAL_PACING].fillna('')
df1_deduplicated[RPT_COL_TOTAL_PACING] = df1_deduplicated[RPT_COL_TOTAL_PACING].apply(lambda x: f"{int(x)}%" if x != '' else x)

##Delivery Status   
##Formula: =iferror(IF(AND(Y3>F3, D3="CPM", P3/AG3<97%), "Underdelivery",
# IF(AND(Y3>F3, D3="CPM", OR(P3/AG3>115%, P3-AG3>50000)), "Overdelivery",
#    IF(AND(Y3>F3, D3="CPM", P3/AG3>=97%, P3/AG3<=115%), "On Target",
#      IF(AND(Y3>F3, D3="MS", O3/AG3<100%), "Underdelivery",
#        IF(AND(Y3>F3, D3="MS", OR(O3/AG3>103%, O3-AG3>500)), "Overdelivery",
#          IF(AND(Y3>F3, D3="MS", O3/AG3>=100%, O3/AG3<=103%), "On Target",
#            IF(AND(Y3>F3, D3="CPV", R3/AG3<100%), "Underdelivery",
#             IF(AND(Y3>F3, D3="CPV", OR(R3/AG3>103%, R3-AG3>50000)), "Overdelivery",
#                IF(AND(Y3>F3, D3="CPV", R3/AG3>=100%, R3/AG3<=103%), "On Target", "")
##Python
def calculate_delivery_status(row):
    try:
        if row[RPT_COL_AUTO_PACING_CYCLE_THRESHOLD] == "Campaign Ended":
            goal = row[RPT_COL_GOAL]
            total_target_spend_per_imprviews = float(str(row[RPT_COL_TOTAL_TARGET_SPEND_PER_IMPRVIEWS]).replace(',', ''))

            if goal == "CPM":
                impressions = float(str(row[RPT_COL_TOTAL_IMPRESSIONS]).replace(',', ''))
                ratio = impressions / total_target_spend_per_imprviews
                if ratio < 0.97:
                    return "Underdelivery"
                elif ratio > 1.15 or (impressions - total_target_spend_per_imprviews > 50000):
                    return "Overdelivery"
                elif 0.97 <= ratio <= 1.15:
                    return "On Target"
            elif goal == "CPV":
                views = float(str(row[RPT_COL_TOTAL_VIEWS]).replace(',', ''))
                ratio = views / total_target_spend_per_imprviews
                if ratio < 1.00:
                    return "Underdelivery"
                elif ratio > 1.03 or (views - total_target_spend_per_imprviews > 50000):
                    return "Overdelivery"
                elif 1.00 <= ratio <= 1.03:
                    return "On Target"
            elif goal == "MS":
                cost = float(str(row[RPT_COL_TOTAL_PUB_COST]).replace(',', ''))
                ratio = cost / total_target_spend_per_imprviews
                if ratio < 1.00:
                    return "Underdelivery"
                elif ratio > 1.03 or (cost - total_target_spend_per_imprviews > 500):  # Corrected to use `cost` instead of `views`
                    return "Overdelivery"
                elif 1.00 <= ratio <= 1.03:
                    return "On Target"
        else:
            return ""
    except Exception as e:
        # Return the exception for troubleshooting
        return str(e)

    return ""

# Apply the function to each row
df1_deduplicated[RPT_COL_DELIVERY_STATUS] = df1_deduplicated.apply(calculate_delivery_status, axis=1)


##Auto. Pacing Cycle Target Remaining (Spend, Impr., Views) 
##Formula: =IF(A2<>"",IF(D2="CPM",G2-T2,if(D2="ms",G2-S2,IF(D2="CPV",G2-V2))),"")
##Python
def calculate_auto_pacing_cycle_target_remaining(row):
    if row[RPT_COL_CAMPAIGN] != "":  # Checks if the campaign column is not empty
        if row[RPT_COL_GOAL] == "CPM":
            return row[RPT_COL_TARGET_IMPR_PER_SPENDVIEWS] - row[RPT_COL_AUTO_PACING_CYCLE_IMPR] 
        elif row[RPT_COL_GOAL] == "MS":
            return row[RPT_COL_TARGET_IMPR_PER_SPENDVIEWS] - row[RPT_COL_AUTO_PACING_CYCLE_PUB_COST]
        elif row[RPT_COL_GOAL] == "CPV":
            return row[RPT_COL_TARGET_IMPR_PER_SPENDVIEWS] - row[RPT_COL_AUTO_PACING_CYCLE_VIEWS]
        else:
            return ""
    else:
        return ""

# Apply the function to each row of the DataFrame
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_TARGET_REMAINING_SPEND_PER_IMPRVIEWS] = df1_deduplicated.apply(calculate_auto_pacing_cycle_target_remaining, axis=1)
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_TARGET_REMAINING_SPEND_PER_IMPRVIEWS] = df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_TARGET_REMAINING_SPEND_PER_IMPRVIEWS].apply(
    lambda x: f'{x:.2f}' if not pd.isna(pd.to_numeric(x, errors='coerce')) else x
)


##Pacing Cycle Daily Allocation (Spend, Impr., Views)   
##Formula: =iferror(IF(D2="MS",(G2-S2)/AB2,IF(D2="CPM",(G2-T2)/AB2,if(D2="CPV",(G2-V2)/AB2,""))))
##Python
def calculate_pacing_cycle_daily_allocation(row):
    try:
        if row[RPT_COL_GOAL] == "MS":
            remaining_target = row[RPT_COL_TARGET_IMPR_PER_SPENDVIEWS] - row[RPT_COL_AUTO_PACING_CYCLE_PUB_COST]
            return remaining_target / row[RPT_COL_AUTO_PACING_CYCLE_DAYS_REMAINING] if row[RPT_COL_AUTO_PACING_CYCLE_DAYS_REMAINING] != 0 else 0
        elif row[RPT_COL_GOAL] == "CPM":
            remaining_target = row[RPT_COL_TARGET_IMPR_PER_SPENDVIEWS] - row[RPT_COL_AUTO_PACING_CYCLE_IMPR]
            return remaining_target / row[RPT_COL_AUTO_PACING_CYCLE_DAYS_REMAINING] if row[RPT_COL_AUTO_PACING_CYCLE_DAYS_REMAINING] != 0 else 0
        elif row[RPT_COL_GOAL] == "CPV":
            remaining_target = row[RPT_COL_TARGET_IMPR_PER_SPENDVIEWS] - row[RPT_COL_AUTO_PACING_CYCLE_VIEWS]
            return remaining_target / row[RPT_COL_AUTO_PACING_CYCLE_DAYS_REMAINING] if row[RPT_COL_AUTO_PACING_CYCLE_DAYS_REMAINING] != 0 else 0
        else:
            return 0  # Return 0 or an appropriate value for rows that do not match any campaign type conditions
    except:
        return 0  # Handles division by zero or any other calculation error gracefully

# Apply the function to each row of the DataFrame
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_DAILY_ALLOCATION_SPEND_PER_IMPRVIEWS] = df1_deduplicated.apply(calculate_pacing_cycle_daily_allocation, axis=1)
df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_DAILY_ALLOCATION_SPEND_PER_IMPRVIEWS] = df1_deduplicated[RPT_COL_AUTO_PACING_CYCLE_DAILY_ALLOCATION_SPEND_PER_IMPRVIEWS].apply(
    lambda x: f'{x:.2f}' if not pd.isna(pd.to_numeric(x, errors='coerce')) else x
)


##Recommended Daily Budget
##Formula: =IFERROR(IF(AND(AA2=0,D2="ms"),AC2,IF(AND(AA2=0,D2="CPM"),(O2/P2)*AC2,IF(AND(AA2=0,D2="cpv"),O2/V2*AC2,IF(D2="MS",(G2-S2)/AB2,IF(D2="CPM",(O2/P2)*AO2,IF(D2="CPV",O2/V2*AO2,"")))))),"")
##Python
def convert_to_float(value):
    """
    Converts a value to float. If the value is a string that includes commas,
    it removes the commas before conversion. If the value is empty or not convertible,
    it returns 0.0.
    
    Parameters:
    value (str or float): The value to convert.
    
    Returns:
    float: The converted value or 0.0 if conversion is not possible.
    """
    try:
        # Check if the value is already a float
        if isinstance(value, float):
            return value
        # If the value is a string, remove commas and convert to float
        if isinstance(value, str) and value:
            return float(value.replace(',', ''))
    except ValueError:
        pass
    # Return 0.0 for empty strings or values that cannot be converted
    return 0.0
def convert_to_float(value):
    """
    Converts a value to float. If the value is a string that includes commas,
    it removes the commas before conversion. If the value is empty or not convertible,
    it returns 0.0.
    
    Parameters:
    value (str or float): The value to convert.
    
    Returns:
    float: The converted value or 0.0 if conversion is not possible.
    """
    try:
        # Check if the value is already a float
        if isinstance(value, float):
            return value
        # If the value is a string, remove commas and convert to float
        if isinstance(value, str) and value:
            return float(value.replace(',', ''))
    except ValueError:
        pass
    # Return 0.0 for empty strings or values that cannot be converted
    return 0.0
def calculate_recommended_daily_budget(row):
    """
    Calculates the recommended daily budget based on campaign goal, remaining targets, and pacing.
    
    Parameters:
    row (pd.Series): A row from the DataFrame.
    
    Returns:
    str: The recommended daily budget for the campaign as a formatted string.
    """
    try:
        # Check if the Auto. Pacing Cycle Threshold indicates the campaign has ended
        if row['Auto. Pacing Cycle Threshold'] == "Campaign Ended":
            return ""
            
        # Use the convert_to_float function for all conversions
        days_elapsed = convert_to_float(row[RPT_COL_AUTO_PACING_CYCLE_DAYS_ELAPSED])
        pub_cost = convert_to_float(row[RPT_COL_TOTAL_PUB_COST])
        total_impressions = convert_to_float(row[RPT_COL_TOTAL_IMPRESSIONS])
        total_views = convert_to_float(row[RPT_COL_AUTO_PACING_CYCLE_VIEWS])
        daily_target = convert_to_float(row[RPT_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS])
        goal = row[RPT_COL_GOAL]
        if days_elapsed == 0 and goal == "MS":
            result = daily_target
        elif days_elapsed == 0 and goal == "CPM":
            result = (pub_cost / total_impressions) * daily_target
        elif days_elapsed == 0 and goal == "CPV":
            result = (pub_cost / total_views) * daily_target    
        elif goal == "MS":
            target = convert_to_float(row[RPT_COL_TARGET_IMPR_PER_SPENDVIEWS])
            remaining_days = convert_to_float(row[RPT_COL_AUTO_PACING_CYCLE_DAYS_REMAINING])
            apc_pub_cost = convert_to_float(row[RPT_COL_AUTO_PACING_CYCLE_PUB_COST])
            result = (target - apc_pub_cost) / remaining_days if remaining_days else ""
        elif goal == "CPM":
            allocation = convert_to_float(row[RPT_COL_AUTO_PACING_CYCLE_DAILY_ALLOCATION_SPEND_PER_IMPRVIEWS])
            result = (pub_cost / total_impressions) * allocation if total_impressions else ""
        elif goal == "CPV":
            allocation = convert_to_float(row[RPT_COL_AUTO_PACING_CYCLE_DAILY_ALLOCATION_SPEND_PER_IMPRVIEWS])
            result = (pub_cost / total_views) * allocation if total_views else ""
        else:
            result = "" 

        if result <= 0:
            result = 1.00  # Set a default value if the result is less than or equal to 0

    except Exception as e:
        print(f"Error calculating recommended daily budget for campaign {row[RPT_COL_CAMPAIGN]}: {e}")
        result = ""  # Set a default value in case of an error

    # Ensure result is a float before formatting
    try:
        result = float(result)
        return f"{result:.2f}"
    except ValueError:
        # If result cannot be converted to float, return it as is
        return result
        
# Apply the adjusted function
df1_deduplicated[RPT_COL_RECOMMENDED_DAILY_BUDGET] = df1_deduplicated.apply(calculate_recommended_daily_budget, axis=1)
#df1_deduplicated[RPT_COL_RECOMMENDED_DAILY_BUDGET] = df1_deduplicated[RPT_COL_RECOMMENDED_DAILY_BUDGET].apply(
#    lambda x: f'{x:.2f}' if isinstance(x, (int, float)) and not pd.isna(x) else ""
#)
# Assuming df1_deduplicated is your final DataFrame after all transformations
print(df1_deduplicated[[RPT_COL_CAMPAIGN, RPT_COL_RECOMMENDED_DAILY_BUDGET]].head())  # Print a sample for verification
outputDf = df1_deduplicated
Post generated on 2024-05-15 07:44:05 GMT
06 Mar 2024
« Script 771: Script Campaign Bulk Sheet Script 773: Script New Campaign Calculations »
MarinOne Scripts Creator's Corner
