DS3000-Project

Member Contribution Plan:
Clean data – Darsheen
Graphs and visualizations – Hayli
Machine learning – Ethan
Final report and presentation – Eric

import requests import pandas as pd api_key = '7a802e204a08789569034837ff203fb7' url_base = 'https://api.stlouisfed.org/fred/series/observations'

def fetch_and_clean_data(series_ids, start_date='2000-01-01', end_date='2024-12-31'): """ pulls macro economic data from the given series ID with specified start and end date and cleans the data

Args:
    series_ids (list): list of macro economic series id to get from the API
    start_date (str) : start date for collecting data in YY-MM-DD format
    end_date (str) : end date for collecting data in YY-MM-DD format
Returns:
    merged_df (DataFrame): a merged dataframe with numerical and categorical data from
                            the series requested. Each column corrosponds to a series ID 
                            with its values
"""
# creating an empty list to store all values
all_data = []
 
# using for loop to iterate through each series ID to get data from API key  
for series_id in series_ids:
    params = {
        'series_id': series_id,
        'api_key': api_key,
        'file_type': 'json',
        'observation_start': start_date,
        'observation_end': end_date
    }
    
    # sending a request to API
    response = requests.get(url_base, params=params)
    data = response.json()['observations']
    
    # Creating the dataframe
    df = pd.DataFrame(data)
    df = df[['date', 'value']]
    
    # converting the values to numeric and date in datetime format
    df['value'] = pd.to_numeric(df['value'], errors='coerce')
    df['date'] = pd.to_datetime(df['date'])
    
    # renaming the value columns to series ID
    df.rename(columns={'value': series_id}, inplace=True)
    
    # Append dataFrame to the list
    all_data.append(df)

# Adding categorical feature: Inflation Level
if 'CPIAUCSL' in series_ids:
    merged_df['Inflation_Level'] = pd.cut(merged_df['CPIAUCSL'], 
                                          bins=[-float('inf'), 2, 3, float('inf')], 
                                          labels=['Low', 'Medium', 'High'])

# calculating GDP growth rate and categorizing it as recession, stagnation or growth
if 'GDP' in series_ids:
    merged_df['GDP_Growth_Rate'] = merged_df['GDP'].pct_change() * 100
    merged_df['GDP_Growth_Stage'] = pd.cut(merged_df['GDP_Growth_Rate'], 
                                           bins=[-float('inf'), 0, 2, float('inf')], 
                                           labels=['Recession', 'Stagnation', 'Growth'])

# Saving the merged DataFrame to a CSV file
merged_df.to_csv('macro_data.csv')

return merged_df

list of series ID

series_ids = ['GDP', 'UNRATE', 'CPIAUCSL', 'RSAFS', 'INDPRO'] merged_df = fetch_and_clean_data(series_ids)

merged_df.head()