seasonality by month SPY x20yrs python

import yfinance as yf
import pandas as pd
import plotly.graph_objects as go

# Fetch SPY data for the past 20 years
spy_data = yf.download('SPY', start='2004-01-01', end='2024-09-08')

# Resample to monthly data (last business day of each month)
spy_monthly = spy_data['Adj Close'].resample('M').ffill()

# Create a DataFrame for percentage changes
monthly_pct_change = spy_monthly.pct_change().dropna() * 100

# Add Year and Month columns for grouping
monthly_pct_change = monthly_pct_change.to_frame(name='Monthly Change')
monthly_pct_change['Year'] = monthly_pct_change.index.year
monthly_pct_change['Month'] = monthly_pct_change.index.strftime('%b')

# Calculate percentage change for each year and month
yearly_changes = monthly_pct_change.groupby(['Year', 'Month'])['Monthly Change'].mean().reset_index()

# Create a pivot table for easier plotting, with months sorted correctly
month_order = ['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']
pivot_table = yearly_changes.pivot(index='Year', columns='Month', values='Monthly Change')[month_order]

# Create a DataFrame to store data for plotting
plot_data = []
plot_data_totals = []

# Loop through each month to prepare data for plotting
for month in month_order:
    # Data for each month across all years
    month_data = pivot_table[month].dropna()
    month_avg = month_data.mean()

    # Add all years' data for the current month
    for year in month_data.index:
        plot_data.append({'Month': month, 'Year': str(year), 'Monthly Change': month_data[year]})

    # Add the total average at the end of the current month's data
    plot_data_totals.append({'Month': month, 'Year': 'Total Avg', 'Monthly Change': month_avg})

# Convert plot data to DataFrame
plot_df = pd.DataFrame(plot_data)
plot_df_totals = pd.DataFrame(plot_data_totals)

# Define consistent colors for each month
colors = {
    'Jan': '#1f77b4', 'Feb': '#ff7f0e', 'Mar': '#2ca02c', 'Apr': '#d62728',
    'May': '#9467bd', 'Jun': '#8c564b', 'Jul': '#e377c2', 'Aug': '#7f7f7f',
    'Sep': '#bcbd22', 'Oct': '#17becf', 'Nov': '#1f77b4', 'Dec': '#ff7f0e'
}

# Plot data organized by Month, grouped correctly
fig = go.Figure()

# Plot each month with grouped years
for month in month_order:
    month_df = plot_df[plot_df['Month'] == month]
    fig.add_trace(go.Bar(x=month_df['Month'] + ' ' + month_df['Year'],
                         y=month_df['Monthly Change'],
                         name=month,
                         marker_color=colors[month]))  # Set color for each month

# Plot each month's total
for month in month_order:
    month_df = plot_df_totals[plot_df_totals['Month'] == month]
    fig.add_trace(go.Bar(x=month_df['Month'] + ' ' + month_df['Year'],
                         y=month_df['Monthly Change'],
                         name=month,
                         marker_color=colors[month],  # Match color for totals
                         showlegend=False))  # Hide legend for total bars

fig.update_layout(
    title="SPY Monthly Percentage Change Over the Past 20 Years (Grouped by Month)",
    xaxis_title="Month and Year",
    yaxis_title="Percentage Change",
    barmode='group'
)

fig.show()