# %% # imports import gradio as gr import matplotlib.pyplot as plt import numpy as np import pandas as pd # %% # import data data = pd.read_excel("graph_data.xlsx", sheet_name = "Sheet1", header = 0 ) # clean data # rename year column data = data.rename(columns = {"Unnamed: 0": "Year"}) # drop rows without data # data = data.drop([0,1,2,3,4,5,6,7,8,9,10,11]) # keep columns of interest data = data[[ "Year", "Total Revenues", "Debt Balance", "Revenues ex SS OASDI", "Average Rate on Federal Debt", "GDP", "Net Interest" ]] print(data) # %% baseline_interest_rate = 3.67 baseline_revenues = 15928.73 def plot_interest_coverage(interest_rate, revenues): # calculate the yearly increase in the interest rate based on the projected interest rate in 2054 interest_rate_yearly_increase = (interest_rate - baseline_interest_rate) / (2054 - 2025) / 100 # calculate the yearly increase in revenues based on the projected interest rate in 2054 revenues_yearly_increase = (revenues - baseline_revenues) / (2054 - 2025) # add a baseline net interest / revenues column data["Net Interest / Revenues (Baseline)"] = data["Net Interest"] / data["Total Revenues"] # add a baseline net interest / revenues ex SS OASDI column data["Net Interest / Revenues ex SS OASDI (Baseline)"] = data["Net Interest"] / data["Revenues ex SS OASDI"] # add a projected average rate on federal debt column data["Average Rate on Federal Debt (Projected)"] = np.where( data["Year"].astype(int) < 2026, data["Average Rate on Federal Debt"], data["Average Rate on Federal Debt"] + (interest_rate_yearly_increase * (data["Year"].astype(int) - 2025))) # add a projected revenues column data["Total Revenues (Projected)"] = np.where( data["Year"].astype(int) < 2026, data["Revenues"], data["Revenues"] + (revenues_yearly_increase * (data["Year"].astype(int) - 2025))) # add a projected revenues ex SS OASDI column data["Revenues ex SS OASDI (Projected)"] = np.where( data["Year"].astype(int) < 2026, data["Revenues ex SS OASDI"], data["Revenues ex SS OASDI"] + (revenues_yearly_increase * (data["Year"].astype(int) - 2025))) # add a projected interest / revenues column data["Net Interest / Revenues (Projected)"] = np.where( data["Year"].astype(int) < 2026, data["Net Interest / Revenues (Baseline)"], data["Average Rate on Federal Debt (Projected)"] * data["Debt Balance"] / data["Total Revenues (Projected)"]) # add a projected interest / revenues ex SS OASDI column data["Net Interest / Revenues ex SS OASDI (Projected)"] = np.where( data["Year"].astype(int) < 2026, data["Net Interest / Revenues ex SS OASDI (Baseline)"], data["Average Rate on Federal Debt (Projected)"] * data["Debt Balance"] / data["Revenues ex SS OASDI (Projected)"]) # Create the plot plt.figure(figsize = (10, 4.8)) # plot average rate on federal debt plt.plot( data["Year"], data["Average Rate on Federal Debt"], color = "Green", label = "Average Rate on Federal Debt" ) # plot average rate on federal debt projected plt.plot( data["Year"], data["Average Rate on Federal Debt (Projected)"], color = "Green", label = "Average Rate on Federal Debt (Projected)", linestyle = "--" ) # plot interest / revenues (baseline) plt.plot( data["Year"], data["Net Interest / Revenues (Baseline)"], color = "Blue", label = "Net Interest / Revenues (Baseline)" ) # plot interest / revenues (projected) plt.plot( data["Year"], data["Net Interest / Revenues (Projected)"], color = "Blue", label = "Net Interest / Revenues (Projected)", linestyle = "--" ) # plot interest / revenues ex ss oasdi (baseline) plt.plot( data["Year"], data["Net Interest / Revenues ex SS OASDI (Baseline)"], color = "Orange", label = "Net Interest / Revenues ex SS OASDI (Baseline)" ) # plot interest / revenues ex ss oasdi (projected) plt.plot( data["Year"], data["Net Interest / Revenues ex SS OASDI (Projected)"], color = "Orange", label = "Net Interest / Revenues ex SS OASDI (Projected)", linestyle = "--" ) plt.title("Interest as Share of Revenues Through 2054") plt.legend(loc = "upper left") plt.ylim(0, 1.05) plt.yticks(np.arange(0, 1.1, 0.1)) plt.xticks(range(1940,2055,4), rotation = 45) plt.axvline(x = 2024, color = "black") plt.grid(True, axis = 'y', linestyle = '--', linewidth = 0.7) # Save the plot to a file plt.savefig("interest_coverage.png") plt.close() # Return the path to the image file return "interest_coverage.png" # %% interest_rate_lowerbound = 0 interest_rate_upperbound = 15 revenues_lower_bound = 10000 revenues_upper_bound = 20000 with gr.Blocks() as interface: # Create the image output graph = gr.Image(type="filepath", label = "Graph", value = plot_interest_coverage(baseline_interest_rate, baseline_revenues)) # Create the slider input below the image for projected interest rate interest_rate_slider = gr.Slider( interest_rate_lowerbound, interest_rate_upperbound, step = .1, value = baseline_interest_rate, label = "2054 Projected Average Interest Rate on Federal Debt" ) # Create the slider input below the image for projected revenues revenues_slider = gr.Slider( revenues_lower_bound, revenues_upper_bound, step = 50, value = baseline_revenues, label = "2054 Projected Revenues ($B)" ) gr.Markdown('

Source: CBO June 2024 An Update to the Budget and Economic Outlook: 2024 to 2034, March 2024 report The Long-Term Budget Outlook: 2024 to 2054, and author\'s calculations. Scenario of higher interest rate is author\'s calculations. Data points for 2035-2054 calculated to build on CBO\'s June 2024 10-year update and be consistent with the March 2024 long-term report. Historical Social Security OASDI payroll tax revenue from Table 4-3 of the SSA\'s Trust Fund Data, and projections from the CBO\'s August 2024 Long Term Projections for Social Security.

') # Set up an action that updates the graph when the interest rate slider value changes interest_rate_slider.change( plot_interest_coverage, inputs = [interest_rate_slider, revenues_slider], outputs = graph ) # Set up an action that updates the graph when the revenues slider value changes revenues_slider.change( plot_interest_coverage, inputs = [interest_rate_slider, revenues_slider], outputs = graph ) # Launch the interface interface.launch(share = True)