Script 765: Pacing New Campaign Calculations

Purpose

The Python script calculates pacing metrics and recommended daily budgets for digital advertising campaigns based on various campaign parameters and goals.
To Elaborate

The script is designed to manage and calculate pacing metrics for digital advertising campaigns, ensuring that they meet their goals within specified timeframes. It processes campaign data to compute key metrics such as total days, days elapsed, and days remaining for each campaign. The script also calculates daily targets and expected values based on campaign goals like CPM (Cost Per Mille), CPV (Cost Per View), and MS (Media Spend). It evaluates the pacing status of campaigns, categorizing them as under-pacing, on pace, or over-pacing, and determines the delivery status as underdelivery, on target, or overdelivery. The script ultimately provides a recommended daily budget to help campaigns stay on track, adjusting for factors like campaign type and pacing cycle.
Walking Through the Code

Initialization and Setup
- The script begins by determining if it is running on a server or locally. If local, it loads a data source dictionary from a pickle file.
- It sets up necessary imports, including pandas and numpy, and configures pandas display settings.
Data Preparation
- The script cleans and prepares the input data by converting date columns to datetime objects and stripping unwanted characters from numeric columns.
- It initializes columns for calculations, such as ‘Auto. Pacing Cycle Start Date’ and ‘Auto. Pacing Cycle End Date’, using custom functions to determine these dates based on campaign parameters.
Metric Calculations
- Functions are defined to calculate various metrics, including total days, days elapsed, and days remaining for each campaign.
- The script computes daily targets and expected values based on campaign goals, adjusting for pacing cycles and campaign types.
Pacing and Delivery Status
- The script evaluates the pacing status of campaigns, determining if they are under-pacing, on pace, or over-pacing.
- It calculates the delivery status, categorizing campaigns as underdelivery, on target, or overdelivery based on their performance against targets.
Recommended Daily Budget
- A function calculates the recommended daily budget for each campaign, considering factors like remaining targets and pacing cycle allocations.
- The script formats and outputs the final results, including the recommended daily budget for each campaign.
Output and Debugging
- The script outputs the final DataFrame with calculated metrics and recommended budgets.
- If running locally, it saves the output and debug data to CSV files for further analysis.
Vitals

Script ID : 765
Client ID / Customer ID: 1306927185 / 60270139
Action Type: Bulk Upload
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, Auto. Pacing Daily Target (Spend/Impressions/Views), Installments, Pacing Calculation Date, Pacing Cycle Daily Allocation (Spend/Impr./Views)
Linked Datasource: M1 Report
Reference Datasource: None
Owner: ascott@marinsoftware.com (ascott@marinsoftware.com)
Created by ascott@marinsoftware.com on 2024-03-06 21:59
Last Updated by ascott@marinsoftware.com on 2024-09-11 18:00
> See it in Action

Python Code

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

########### START - Local Mode Config ###########
# Step 1: Uncomment download_preview_input flag and run Preview successfully with the Datasources you want
download_preview_input=False
# Step 2: In MarinOne, go to Scripts -> Preview -> Logs, download 'dataSourceDict' pickle file, and update pickle_path below
# pickle_path = ''
pickle_path = '/Users/mhuang/Downloads/pickle/mary_beth_pacing_new_calc_20240905.pkl'
# Step 3: Copy this script into local IDE with Python virtual env loaded with pandas and numpy.
# Step 4: Run locally with below code to init dataSourceDict

# determine if code is running on server or locally
def is_executing_on_server():
    try:
        # Attempt to access a known restricted builtin
        dict_items = dataSourceDict.items()
        return True
    except NameError:
        # NameError: dataSourceDict object is missing (indicating not on server)
        return False

local_dev = False

if is_executing_on_server():
    print("Code is executing on server. Skip init.")
elif len(pickle_path) > 3:
    print("Code is NOT executing on server. Doing init.")
    local_dev = True
    # load dataSourceDict via pickled file
    import pickle
    dataSourceDict = pickle.load(open(pickle_path, 'rb'))

    # print shape and first 5 rows for each entry in dataSourceDict
    for key, value in dataSourceDict.items():
        print(f"Shape of dataSourceDict[{key}]: {value.shape}")
        # print(f"First 5 rows of dataSourceDict[{key}]:\n{value.head(5)}")

    # set outputDf same as inputDf
    inputDf = dataSourceDict["1"]
    outputDf = inputDf.copy()

    # setup timezone
    import datetime
    # Chicago Timezone is GMT-5. Adjust as needed.
    CLIENT_TIMEZONE = datetime.timezone(datetime.timedelta(hours=-5))

    # import pandas
    import pandas as pd
    import numpy as np

    # other imports
    import re
    import urllib

    # import Marin util functions
    # from marin_scripts_utils import tableize, select_changed

    # pandas settings
    pd.set_option('display.max_columns', None)  # Display all columns
    pd.set_option('display.max_colwidth', None)  # Display full content of each column

else:
    print("Running locally but no pickle path defined. dataSourceDict not loaded.")
    exit(1)
########### END - Local Mode Setup ###########


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'
RPT_COL_INSTALLMENTS = 'Installments'
RPT_COL_PACING_CYCLE_DAILY_ALLOCATION_SPEND_IMPR_VIEWS = 'Pacing Cycle Daily Allocation (Spend/Impr./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_INSTALLMENTS = 'Installments'
BULK_COL_PACING_CALCULATION_DATE = 'Pacing Calculation Date'
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_PACING_CYCLE_DAILY_ALLOCATION_SPEND_IMPR_VIEWS = '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'

print("inputDf.shape", inputDf.shape)
print("inputDf.dtypes\n", inputDf.dtypes)

### Data Cleanup and Preliminaries
# Strip dollar signs and commas, then convert to numeric
spendviews = inputDf[RPT_COL_TARGET_IMPR_PER_SPENDVIEWS].astype(str).str.replace('$', '').str.replace(',', '')
inputDf[RPT_COL_TARGET_IMPR_PER_SPENDVIEWS] = pd.to_numeric(spendviews, errors='coerce')

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

# Today
inputDf[RPT_COL_TODAYS_DATE] = pd.to_datetime(today)

# Pacing Calculation Date
inputDf[BULK_COL_PACING_CALCULATION_DATE] = pd.to_datetime(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

def calculate_auto_pacing_cycle_start_date(row):
    if pd.notna(row['Campaign']):
        if row['Installments'] == 'Manual':
            return row['Auto. Pacing Cycle Start Date']
        elif row['Installments'] == 'Auto':
            if row['Todays Date'] > row['Pacing - End Date']:
                return ""
            elif (row['Pacing - Start Date'].year == row['Todays Date'].year and
                  row['Pacing - Start Date'].month == row['Todays Date'].month):
                return row['Pacing - Start Date']
            elif row['Total Days'] <= 31:
                return row['Pacing - Start Date']
            else:
                return (row['Todays Date'] - pd.offsets.MonthBegin(1)).strftime('%Y-%m-%d')
        elif row['Installments'] == 'Off' or row['Installments'] == "":
            return row['Pacing - Start Date']
        else:
            return row['Pacing - Start Date']
    else:
        return ""

# Apply the function to each row of the DataFrame
inputDf[RPT_COL_AUTO_PACING_CYCLE_START_DATE] = inputDf.apply(calculate_auto_pacing_cycle_start_date, axis=1)
inputDf[RPT_COL_AUTO_PACING_CYCLE_START_DATE] = pd.to_datetime(inputDf[RPT_COL_AUTO_PACING_CYCLE_START_DATE], errors='coerce')

##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

def calculate_auto_pacing_cycle_end_date(row):
    if pd.notna(row['Campaign']):
        if row['Installments'] == 'Manual':
            return row['Auto. Pacing Cycle End Date']
        elif row['Installments'] == 'Auto':
            if row['Todays Date'] > row['Pacing - End Date']:
                return ""
            elif (row['Pacing - End Date'].year == row['Todays Date'].year and
                  row['Pacing - End Date'].month == row['Todays Date'].month):
                return row['Pacing - End Date']
            elif row['Total Days'] <= 31:
                return row['Pacing - End Date']
            else:
                return (row['Todays Date'] + pd.offsets.MonthEnd(0)).strftime('%Y-%m-%d')
        elif row['Installments'] == 'Off' or row['Installments'] == "":
            return row['Pacing - End Date']
        else:
            return row['Pacing - End Date']
    else:
        return ""

# Apply the function to each row of the DataFrame
inputDf[RPT_COL_AUTO_PACING_CYCLE_END_DATE] = inputDf.apply(calculate_auto_pacing_cycle_end_date, axis=1)
inputDf[RPT_COL_AUTO_PACING_CYCLE_END_DATE] = pd.to_datetime(inputDf[RPT_COL_AUTO_PACING_CYCLE_END_DATE], errors='coerce')


##Total Days    
##Formula: =if(A2<>"",IF(F2-E2+1<1,"",F2-E2+1),"")
##Python 

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
inputDf[RPT_COL_TOTAL_DAYS] = inputDf.apply(calculate_total_days, axis=1)

#### calculate metric totals for Pacing and Auto Pacing date ranges

## Make copies of metric values before zeroing out according to date range
# List of columns to be assigned the same value
columns_to_assign = [
    [RPT_COL_TOTAL_IMPRESSIONS, RPT_COL_PACING_CYCLE_IMPRESSIONS, RPT_COL_AUTO_PACING_CYCLE_IMPR],
    [RPT_COL_TOTAL_CLICKS, RPT_COL_PACING_CYCLE_CLICKS, RPT_COL_AUTO_PACING_CYCLE_CLICKS],
    [RPT_COL_TOTAL_PUB_COST, RPT_COL_PACING_CYCLE_PUB_COST, RPT_COL_AUTO_PACING_CYCLE_PUB_COST],
    [RPT_COL_TOTAL_VIEWS, RPT_COL_PACING_CYCLE_VIEWS, RPT_COL_AUTO_PACING_CYCLE_VIEWS]
]

# Corresponding values to assign
values_to_assign = [RPT_COL_IMPR, RPT_COL_CLICKS, RPT_COL_PUB_COST, RPT_COL_VIDEO_VIEWS]

# Assign values to columns
for cols, value in zip(columns_to_assign, values_to_assign):
    for col in cols:
        inputDf[col] = inputDf[value]

## Zero out values if Date not within PACING_START/END or AUTO_PACING_START/END
def zero_out_values(row):
    if not (row[RPT_COL_PACING__START_DATE] <= row[RPT_COL_DATE] <= row[RPT_COL_PACING__END_DATE]):
        row[RPT_COL_PACING_CYCLE_IMPRESSIONS] = 0
        row[RPT_COL_PACING_CYCLE_CLICKS] = 0
        row[RPT_COL_PACING_CYCLE_PUB_COST] = 0
        row[RPT_COL_PACING_CYCLE_VIEWS] = 0

    if not (row[RPT_COL_AUTO_PACING_CYCLE_START_DATE] <= row[RPT_COL_DATE] <= row[RPT_COL_AUTO_PACING_CYCLE_END_DATE]):
        row[RPT_COL_AUTO_PACING_CYCLE_IMPR] = 0
        row[RPT_COL_AUTO_PACING_CYCLE_CLICKS] = 0
        row[RPT_COL_AUTO_PACING_CYCLE_PUB_COST] = 0
        row[RPT_COL_AUTO_PACING_CYCLE_VIEWS] = 0

    return row

# Apply the zero_out_values function to each row
inputDf = inputDf.apply(zero_out_values, axis=1)

# Now agg to get subtotal for each date range
agg_dict = {
    RPT_COL_TOTAL_IMPRESSIONS: 'sum',
    RPT_COL_TOTAL_CLICKS: 'sum',
    RPT_COL_TOTAL_PUB_COST: 'sum',
    RPT_COL_TOTAL_VIEWS: 'sum',
    RPT_COL_PACING_CYCLE_IMPRESSIONS: 'sum',
    RPT_COL_PACING_CYCLE_CLICKS: 'sum',
    RPT_COL_PACING_CYCLE_PUB_COST: 'sum',
    RPT_COL_PACING_CYCLE_VIEWS: 'sum',
    RPT_COL_AUTO_PACING_CYCLE_PUB_COST: 'sum',
    RPT_COL_AUTO_PACING_CYCLE_IMPR: 'sum',
    RPT_COL_AUTO_PACING_CYCLE_CLICKS: 'sum',
    RPT_COL_AUTO_PACING_CYCLE_VIEWS: 'sum',
}

# Keep the last value for all columns in inputDf that do not appear in agg_dict, excluding RPT_COL_DATE, RPT_COL_ACCOUNT, and RPT_COL_CAMPAIGN
for col in inputDf.columns:
    if col not in agg_dict and col not in [RPT_COL_DATE, RPT_COL_ACCOUNT, RPT_COL_CAMPAIGN]:
        agg_dict[col] = 'last'

debugDf = inputDf

df_campaign_with_totals = inputDf.groupby([RPT_COL_ACCOUNT, RPT_COL_CAMPAIGN]).agg(agg_dict).reset_index()

print("df_campaign_with_totals.shape", df_campaign_with_totals.shape)
print("df_campaign_with_totals.dtypes\n", df_campaign_with_totals.dtypes)

##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:

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
df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_DAYS] = df_campaign_with_totals.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: 

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
df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_DAYS_ELAPSED] = df_campaign_with_totals.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
df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_DAYS_REMAINING] = df_campaign_with_totals.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
df_campaign_with_totals[RPT_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS] = df_campaign_with_totals.apply(calculate_auto_pacing_daily_target, axis=1)
df_campaign_with_totals[RPT_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS] = df_campaign_with_totals[RPT_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS].round(2)
df_campaign_with_totals[RPT_COL_AUTO_PACING_DAILY_TARGET_SPEND_PER_IMPRESSIONSVIEWS] = df_campaign_with_totals[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
df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE] = df_campaign_with_totals.apply(calculate_apc_etd, axis=1)
df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_EXPECTED_TO_DATE] = df_campaign_with_totals[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
df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_PACING] = df_campaign_with_totals.apply(calculate_auto_pacing_cycle_pacing, axis=1)
df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_PACING] = pd.to_numeric(df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_PACING], errors='coerce') * 100.0
df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_PACING] = df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_PACING].round(2)
df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_PACING] = df_campaign_with_totals[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

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
df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_THRESHOLD] = df_campaign_with_totals.apply(calculate_auto_pacing_cycle_threshold, axis=1)


##Total Target (Spend/Impr./Views) v2
##Formula: =if(A22<>"",if(Or(N22="Auto",N22="No",N22=""),IF(F22-E22+1<=31,G22,(G22/30)*(F22-E22)+1),G22),"")
##Python
# Extract and parse target values
df_campaign_with_totals['Target (Impr/Spend/Views)'] = df_campaign_with_totals['Target (Impr/Spend/Views)'].astype(str).replace({',': '', '\s+': ''}, regex=True)
df_campaign_with_totals['Target (Impr/Spend/Views)'] = pd.to_numeric(df_campaign_with_totals['Target (Impr/Spend/Views)'], errors='coerce')
df_campaign_with_totals['Total Days'] = pd.to_numeric(df_campaign_with_totals['Total Days'], errors='coerce')

# Calculate days_diff
df_campaign_with_totals['days_diff'] = (pd.to_datetime(df_campaign_with_totals['Pacing - End Date']) - pd.to_datetime(df_campaign_with_totals['Pacing - Start Date'])).dt.days + 1

# Correct calculation for Total Target (Spend/Impr/Views) dynamically
def calculate_total_target(row):
    if pd.notna(row['Campaign']):
        if row['Installments'] in ["Auto", "No", ""]:
            days_diff = row['days_diff']
            initial_target = row['Target (Impr/Spend/Views)']
            if days_diff <= 31:
                return initial_target
            else:
                return (initial_target / 30) * (days_diff - 1)
        else:
            return row['Target (Impr/Spend/Views)']
    return ""

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

##Total Days    
##Formula: =if(A2<>"",IF(F2-E2+1<1,"",F2-E2+1),"")
##Python 

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
df_campaign_with_totals[RPT_COL_TOTAL_DAYS] = df_campaign_with_totals.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

# 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
df_campaign_with_totals[RPT_COL_TOTAL_DAYS_ELAPSED] = df_campaign_with_totals.apply(calculate_days_elapsed, axis=1)


##Total Daily Target v2   
##Formula: =iferror(AL2/AM2,"")
##Python 
# Verify calculation for Total Daily Target
def safe_division(numerator, denominator):
    try:
        numerator = pd.to_numeric(numerator, errors='coerce')
        if isinstance(denominator, (int, float)) and denominator != 0:
            return numerator / denominator
        else:
            return 0
    except ZeroDivisionError:
        return 0

df_campaign_with_totals[RPT_COL_TOTAL_DAILY_TARGET]  = df_campaign_with_totals.apply(lambda row: safe_division(row[RPT_COL_TOTAL_TARGET_SPEND_PER_IMPRVIEWS], row[RPT_COL_TOTAL_DAYS]), axis=1)
df_campaign_with_totals[RPT_COL_TOTAL_DAILY_TARGET] = df_campaign_with_totals[RPT_COL_TOTAL_DAILY_TARGET].fillna(0).round(2)  # Adjust rounding


##Total Expected to Date v2    
##Formula: =if(AI2<>"",AI2*AJ2,"")
##Python 
# Verify calculation for Total Expected to Date
#current_date = pd.to_datetime("2024-07-25")
#df1_deduplicated[RPT_COL_TOTAL_DAYS_ELAPSED] = (current_date - pd.to_datetime(df1_deduplicated['Pacing - Start Date'])).dt.days

def calculate_expected_to_date(row):
    total_days_elapsed = pd.to_numeric(row['Total Days Elapsed'], errors='coerce')
    total_daily_target = pd.to_numeric(row['Total Daily Target'], errors='coerce')
    if pd.isnull(total_days_elapsed) or pd.isnull(total_daily_target):
        return 0
    return total_days_elapsed * total_daily_target

# Apply the function to each row in the DataFrame
df_campaign_with_totals[RPT_COL_TOTAL_EXPECTED_TO_DATE] = df_campaign_with_totals.apply(calculate_expected_to_date, axis=1)
df_campaign_with_totals[RPT_COL_TOTAL_EXPECTED_TO_DATE] = df_campaign_with_totals[RPT_COL_TOTAL_EXPECTED_TO_DATE].fillna(0).round(2)  # Adjust rounding


##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
df_campaign_with_totals[RPT_COL_TOTAL_PACING] = df_campaign_with_totals.apply(calculate_total_pacing, axis=1)
df_campaign_with_totals[RPT_COL_TOTAL_PACING] = pd.to_numeric(df_campaign_with_totals[RPT_COL_TOTAL_PACING], errors='coerce') * 100.0
df_campaign_with_totals[RPT_COL_TOTAL_PACING] = df_campaign_with_totals[RPT_COL_TOTAL_PACING].round(2)
df_campaign_with_totals[RPT_COL_TOTAL_PACING] = df_campaign_with_totals[RPT_COL_TOTAL_PACING].fillna('')
df_campaign_with_totals[RPT_COL_TOTAL_PACING] = df_campaign_with_totals[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
df_campaign_with_totals[RPT_COL_DELIVERY_STATUS] = df_campaign_with_totals.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
df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_TARGET_REMAINING_SPEND_PER_IMPRVIEWS] = df_campaign_with_totals.apply(calculate_auto_pacing_cycle_target_remaining, axis=1)
df_campaign_with_totals[RPT_COL_AUTO_PACING_CYCLE_TARGET_REMAINING_SPEND_PER_IMPRVIEWS] = df_campaign_with_totals[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
df_campaign_with_totals[RPT_COL_PACING_CYCLE_DAILY_ALLOCATION_SPEND_IMPR_VIEWS] = df_campaign_with_totals.apply(calculate_pacing_cycle_daily_allocation, axis=1)
df_campaign_with_totals[RPT_COL_PACING_CYCLE_DAILY_ALLOCATION_SPEND_IMPR_VIEWS] = df_campaign_with_totals[RPT_COL_PACING_CYCLE_DAILY_ALLOCATION_SPEND_IMPR_VIEWS].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:
        # Calculate the pacing cycle daily allocation and print it
        pacing_cycle_daily_allocation = calculate_pacing_cycle_daily_allocation(row)
        print(f"Pacing Cycle Daily Allocation (Spend/Impr./Views): {pacing_cycle_daily_allocation}")

        if row['Auto. Pacing Cycle Days Elapsed'] == 0:
            if row['Goal'] == "ms":
                return row['Auto. Pacing Daily Target (Spend/Impressions/Views)']
            elif row['Goal'] == "CPM":
                return (row['Total Pub Cost'] / row['Total Impressions']) * row['Auto. Pacing Daily Target (Spend/Impressions/Views)']
            elif row['Goal'] == "cpv":
                return (row['Total Pub Cost'] / row['Auto. Pacing Cycle Views']) * row['Auto. Pacing Daily Target (Spend/Impressions/Views)']
        if row['Goal'] == "MS":
            return (row['Target (Impr/Spend/Views)'] - row['Auto. Pacing Cycle Pub. Cost']) / row['Auto. Pacing Cycle Days Remaining']
        elif row['Goal'] == "CPM":
            return (row['Total Pub Cost'] / row['Total Impressions']) * pacing_cycle_daily_allocation
        elif row['Goal'] == "CPV":
            return (row['Total Pub Cost'] / row['Auto. Pacing Cycle Views']) * pacing_cycle_daily_allocation
        return ""

        if result <= 0:
            result = 0.10  # 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
df_campaign_with_totals[RPT_COL_RECOMMENDED_DAILY_BUDGET] = df_campaign_with_totals.apply(calculate_recommended_daily_budget, axis=1)
df_campaign_with_totals[RPT_COL_RECOMMENDED_DAILY_BUDGET] = df_campaign_with_totals[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(df_campaign_with_totals[[RPT_COL_CAMPAIGN, RPT_COL_RECOMMENDED_DAILY_BUDGET]].head())  # Print a sample for verification
outputDf = df_campaign_with_totals

print("outputDf shape", outputDf.shape)

## local debug
if local_dev:
    output_filename = 'outputDf.csv'
    outputDf.to_csv(output_filename, index=False)
    print(f"Local Dev: Output written to: {output_filename}")

    debug_filename = 'debugDf.csv'
    debugDf.to_csv(debug_filename, index=False)
    print(f"Local Dev: Debug written to: {debug_filename}")## Infinite Digital Pacing Script
Post generated on 2024-11-27 06:58:46 GMT
06 Mar 2024
« Script 763: Script Campaign Bulk Sheet Script 765: Script New Campaign Calculations »
MarinOne Scripts Creator's Corner
