Daily Cost of EC2 Instance Types

import pandas as pd from datetime import datetime, timedelta, timezone if df is not None: # Convert 'lineItem/UsageStartDate' to datetime df['lineItem/UsageStartDate'] = pd.to_datetime(df['lineItem/UsageStartDate']) # Filter out negative 'lineItem/UnblendedCost' df = df[df['lineItem/UnblendedCost'] >= 0] # Filter rows for the last N days cutoff_date = datetime.utcnow().replace(tzinfo=timezone.utc) - timedelta(days=int(last_n_days)) last_n_days_df = df[df['lineItem/UsageStartDate'] > cutoff_date].copy() # Further filter data based on product family and pricing unit condition = (last_n_days_df['product/productFamily'] == 'Compute Instance') & (last_n_days_df['pricing/unit'].isin(['Hours', 'Hrs'])) filtered_df = last_n_days_df[condition].copy() # Group by 'lineItem/UsageStartDate' and 'product/instanceType', then sum 'lineItem/UsageAmount' and 'lineItem/UnblendedCost' result = filtered_df.groupby([filtered_df['lineItem/UsageStartDate'].dt.date, 'product/instanceType']).agg( usage_hours=('lineItem/UsageAmount', 'sum'), usage_cost=('lineItem/UnblendedCost', 'sum') ).reset_index() # Set the properties for your plot context.plot.xlabel = 'Date' context.plot.ylabel = 'EC2 Usage Cost($)' context.plot.title = f'EC2 Cost Usage (Last {last_n_days} Days)' # Loop through each unique instance type and add a trace for each for instance_type in result['product/instanceType'].unique(): instance_data = result[result['product/instanceType'] == instance_type] x = instance_data['lineItem/UsageStartDate'].tolist() # Date on x-axis y = instance_data['usage_cost'].tolist() # Usage cost on y-axis context.plot.add_trace(name=f"EC2- {instance_type}", xpts=x, ypts=y, tracetype="line") #print("Analysis complete.") else: print("Failed to fetch data. Exiting.")