FREDML / src /visualization /chart_generator.py
Edwin Salguero
Prepare for Streamlit Cloud deployment - Add deployment files, fix clustering chart error, update requirements
6ce20d9
#!/usr/bin/env python3
"""
Chart Generator for FRED ML
Creates comprehensive economic visualizations and stores them in S3
"""
import io
import json
import os
from datetime import datetime
from typing import Dict, List, Optional, Tuple
import boto3
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import plotly.express as px
import plotly.graph_objects as go
import seaborn as sns
from plotly.subplots import make_subplots
from sklearn.decomposition import PCA
from sklearn.preprocessing import StandardScaler
# Use hardcoded defaults to avoid import issues
DEFAULT_REGION = 'us-east-1'
# Set style for matplotlib
plt.style.use('seaborn-v0_8')
sns.set_palette("husl")
class ChartGenerator:
"""Generate comprehensive economic visualizations"""
def __init__(self, s3_bucket: str = 'fredmlv1', aws_region: str = None):
self.s3_bucket = s3_bucket
if aws_region is None:
aws_region = DEFAULT_REGION
self.s3_client = boto3.client('s3', region_name=aws_region)
self.chart_paths = []
def create_time_series_chart(self, df: pd.DataFrame, title: str = "Economic Indicators") -> str:
"""Create time series chart and upload to S3"""
try:
fig, ax = plt.subplots(figsize=(15, 8))
for column in df.columns:
if column != 'Date':
ax.plot(df.index, df[column], label=column, linewidth=2)
ax.set_title(title, fontsize=16, fontweight='bold')
ax.set_xlabel('Date', fontsize=12)
ax.set_ylabel('Value', fontsize=12)
ax.legend(fontsize=10)
ax.grid(True, alpha=0.3)
plt.xticks(rotation=45)
plt.tight_layout()
# Save to bytes
img_buffer = io.BytesIO()
plt.savefig(img_buffer, format='png', dpi=300, bbox_inches='tight')
img_buffer.seek(0)
# Upload to S3
chart_key = f"visualizations/time_series_{datetime.now().strftime('%Y%m%d_%H%M%S')}.png"
self.s3_client.put_object(
Bucket=self.s3_bucket,
Key=chart_key,
Body=img_buffer.getvalue(),
ContentType='image/png'
)
plt.close()
self.chart_paths.append(chart_key)
return chart_key
except Exception as e:
print(f"Error creating time series chart: {e}")
return None
def create_correlation_heatmap(self, df: pd.DataFrame) -> str:
"""Create correlation heatmap and upload to S3"""
try:
corr_matrix = df.corr()
fig, ax = plt.subplots(figsize=(12, 10))
sns.heatmap(corr_matrix, annot=True, cmap='coolwarm', center=0,
square=True, linewidths=0.5, cbar_kws={"shrink": .8})
plt.title('Economic Indicators Correlation Matrix', fontsize=16, fontweight='bold')
plt.tight_layout()
# Save to bytes
img_buffer = io.BytesIO()
plt.savefig(img_buffer, format='png', dpi=300, bbox_inches='tight')
img_buffer.seek(0)
# Upload to S3
chart_key = f"visualizations/correlation_heatmap_{datetime.now().strftime('%Y%m%d_%H%M%S')}.png"
self.s3_client.put_object(
Bucket=self.s3_bucket,
Key=chart_key,
Body=img_buffer.getvalue(),
ContentType='image/png'
)
plt.close()
self.chart_paths.append(chart_key)
return chart_key
except Exception as e:
print(f"Error creating correlation heatmap: {e}")
return None
def create_distribution_charts(self, df: pd.DataFrame) -> List[str]:
"""Create distribution charts for each indicator"""
chart_keys = []
try:
for column in df.columns:
if column != 'Date':
fig, ax = plt.subplots(figsize=(10, 6))
# Histogram with KDE
sns.histplot(df[column].dropna(), kde=True, ax=ax)
ax.set_title(f'Distribution of {column}', fontsize=14, fontweight='bold')
ax.set_xlabel(column, fontsize=12)
ax.set_ylabel('Frequency', fontsize=12)
plt.tight_layout()
# Save to bytes
img_buffer = io.BytesIO()
plt.savefig(img_buffer, format='png', dpi=300, bbox_inches='tight')
img_buffer.seek(0)
# Upload to S3
chart_key = f"visualizations/distribution_{column}_{datetime.now().strftime('%Y%m%d_%H%M%S')}.png"
self.s3_client.put_object(
Bucket=self.s3_bucket,
Key=chart_key,
Body=img_buffer.getvalue(),
ContentType='image/png'
)
plt.close()
chart_keys.append(chart_key)
self.chart_paths.append(chart_key)
return chart_keys
except Exception as e:
print(f"Error creating distribution charts: {e}")
return []
def create_pca_visualization(self, df: pd.DataFrame, n_components: int = 2) -> str:
"""Create PCA visualization and upload to S3"""
try:
# Prepare data
df_clean = df.dropna()
scaler = StandardScaler()
scaled_data = scaler.fit_transform(df_clean)
# Perform PCA
pca = PCA(n_components=n_components)
pca_result = pca.fit_transform(scaled_data)
# Create visualization
fig, ax = plt.subplots(figsize=(12, 8))
if n_components == 2:
scatter = ax.scatter(pca_result[:, 0], pca_result[:, 1], alpha=0.6)
ax.set_xlabel(f'PC1 ({pca.explained_variance_ratio_[0]:.1%} variance)', fontsize=12)
ax.set_ylabel(f'PC2 ({pca.explained_variance_ratio_[1]:.1%} variance)', fontsize=12)
else:
# For 3D or more, show first two components
scatter = ax.scatter(pca_result[:, 0], pca_result[:, 1], alpha=0.6)
ax.set_xlabel(f'PC1 ({pca.explained_variance_ratio_[0]:.1%} variance)', fontsize=12)
ax.set_ylabel(f'PC2 ({pca.explained_variance_ratio_[1]:.1%} variance)', fontsize=12)
ax.set_title('PCA Visualization of Economic Indicators', fontsize=16, fontweight='bold')
ax.grid(True, alpha=0.3)
plt.tight_layout()
# Save to bytes
img_buffer = io.BytesIO()
plt.savefig(img_buffer, format='png', dpi=300, bbox_inches='tight')
img_buffer.seek(0)
# Upload to S3
chart_key = f"visualizations/pca_visualization_{datetime.now().strftime('%Y%m%d_%H%M%S')}.png"
self.s3_client.put_object(
Bucket=self.s3_bucket,
Key=chart_key,
Body=img_buffer.getvalue(),
ContentType='image/png'
)
plt.close()
self.chart_paths.append(chart_key)
return chart_key
except Exception as e:
print(f"Error creating PCA visualization: {e}")
return None
def create_forecast_chart(self, historical_data: pd.Series, forecast_data: List[float],
title: str = "Economic Forecast") -> str:
"""Create forecast chart and upload to S3"""
try:
fig, ax = plt.subplots(figsize=(15, 8))
# Plot historical data
ax.plot(historical_data.index, historical_data.values,
label='Historical', linewidth=2, color='blue')
# Plot forecast
forecast_index = pd.date_range(
start=historical_data.index[-1] + pd.DateOffset(months=1),
periods=len(forecast_data),
freq='M'
)
ax.plot(forecast_index, forecast_data,
label='Forecast', linewidth=2, color='red', linestyle='--')
ax.set_title(title, fontsize=16, fontweight='bold')
ax.set_xlabel('Date', fontsize=12)
ax.set_ylabel('Value', fontsize=12)
ax.legend(fontsize=12)
ax.grid(True, alpha=0.3)
plt.xticks(rotation=45)
plt.tight_layout()
# Save to bytes
img_buffer = io.BytesIO()
plt.savefig(img_buffer, format='png', dpi=300, bbox_inches='tight')
img_buffer.seek(0)
# Upload to S3
chart_key = f"visualizations/forecast_{datetime.now().strftime('%Y%m%d_%H%M%S')}.png"
self.s3_client.put_object(
Bucket=self.s3_bucket,
Key=chart_key,
Body=img_buffer.getvalue(),
ContentType='image/png'
)
plt.close()
self.chart_paths.append(chart_key)
return chart_key
except Exception as e:
print(f"Error creating forecast chart: {e}")
return None
def create_regression_diagnostics(self, y_true: List[float], y_pred: List[float],
residuals: List[float]) -> str:
"""Create regression diagnostics chart and upload to S3"""
try:
fig, axes = plt.subplots(2, 2, figsize=(15, 12))
# Actual vs Predicted
axes[0, 0].scatter(y_true, y_pred, alpha=0.6)
axes[0, 0].plot([min(y_true), max(y_true)], [min(y_true), max(y_true)], 'r--', lw=2)
axes[0, 0].set_xlabel('Actual Values')
axes[0, 0].set_ylabel('Predicted Values')
axes[0, 0].set_title('Actual vs Predicted')
axes[0, 0].grid(True, alpha=0.3)
# Residuals vs Predicted
axes[0, 1].scatter(y_pred, residuals, alpha=0.6)
axes[0, 1].axhline(y=0, color='r', linestyle='--')
axes[0, 1].set_xlabel('Predicted Values')
axes[0, 1].set_ylabel('Residuals')
axes[0, 1].set_title('Residuals vs Predicted')
axes[0, 1].grid(True, alpha=0.3)
# Residuals histogram
axes[1, 0].hist(residuals, bins=20, alpha=0.7, edgecolor='black')
axes[1, 0].set_xlabel('Residuals')
axes[1, 0].set_ylabel('Frequency')
axes[1, 0].set_title('Residuals Distribution')
axes[1, 0].grid(True, alpha=0.3)
# Q-Q plot
from scipy import stats
stats.probplot(residuals, dist="norm", plot=axes[1, 1])
axes[1, 1].set_title('Q-Q Plot of Residuals')
axes[1, 1].grid(True, alpha=0.3)
plt.tight_layout()
# Save to bytes
img_buffer = io.BytesIO()
plt.savefig(img_buffer, format='png', dpi=300, bbox_inches='tight')
img_buffer.seek(0)
# Upload to S3
chart_key = f"visualizations/regression_diagnostics_{datetime.now().strftime('%Y%m%d_%H%M%S')}.png"
self.s3_client.put_object(
Bucket=self.s3_bucket,
Key=chart_key,
Body=img_buffer.getvalue(),
ContentType='image/png'
)
plt.close()
self.chart_paths.append(chart_key)
return chart_key
except Exception as e:
print(f"Error creating regression diagnostics: {e}")
return None
def create_clustering_chart(self, df: pd.DataFrame, n_clusters: int = 3) -> str:
"""Create clustering visualization and upload to S3"""
try:
from sklearn.cluster import KMeans
# Prepare data
df_clean = df.dropna()
scaler = StandardScaler()
scaled_data = scaler.fit_transform(df_clean)
# Perform clustering
kmeans = KMeans(n_clusters=n_clusters, random_state=42)
clusters = kmeans.fit_predict(scaled_data)
# PCA for visualization
pca = PCA(n_components=2)
pca_result = pca.fit_transform(scaled_data)
# Create visualization
fig, ax = plt.subplots(figsize=(12, 8))
scatter = ax.scatter(pca_result[:, 0], pca_result[:, 1],
c=clusters, cmap='viridis', alpha=0.6)
# Add cluster centers
centers_pca = pca.transform(kmeans.cluster_centers_)
ax.scatter(centers_pca[:, 0], centers_pca[:, 1],
c='red', marker='x', s=200, linewidths=3, label='Cluster Centers')
ax.set_title(f'K-Means Clustering (k={n_clusters})', fontsize=16, fontweight='bold')
ax.set_xlabel(f'PC1 ({pca.explained_variance_ratio_[0]:.1%} variance)', fontsize=12)
ax.set_ylabel(f'PC2 ({pca.explained_variance_ratio_[1]:.1%} variance)', fontsize=12)
ax.legend()
ax.grid(True, alpha=0.3)
plt.tight_layout()
# Save to bytes
img_buffer = io.BytesIO()
plt.savefig(img_buffer, format='png', dpi=300, bbox_inches='tight')
img_buffer.seek(0)
# Upload to S3
chart_key = f"visualizations/clustering_{datetime.now().strftime('%Y%m%d_%H%M%S')}.png"
self.s3_client.put_object(
Bucket=self.s3_bucket,
Key=chart_key,
Body=img_buffer.getvalue(),
ContentType='image/png'
)
plt.close()
self.chart_paths.append(chart_key)
return chart_key
except Exception as e:
print(f"Error creating clustering chart: {e}")
return None
def generate_comprehensive_visualizations(self, df: pd.DataFrame, analysis_type: str = "comprehensive") -> Dict[str, str]:
"""Generate comprehensive visualizations based on analysis type"""
visualizations = {}
try:
# Always create time series and correlation charts
visualizations['time_series'] = self.create_time_series_chart(df)
visualizations['correlation'] = self.create_correlation_heatmap(df)
visualizations['distributions'] = self.create_distribution_charts(df)
if analysis_type in ["comprehensive", "statistical"]:
# Add PCA visualization
visualizations['pca'] = self.create_pca_visualization(df)
# Add clustering
visualizations['clustering'] = self.create_clustering_chart(df)
if analysis_type in ["comprehensive", "forecasting"]:
# Add forecast visualization (using sample data)
sample_series = df.iloc[:, 0] if not df.empty else pd.Series([1, 2, 3, 4, 5])
sample_forecast = [sample_series.iloc[-1] * 1.02, sample_series.iloc[-1] * 1.04]
visualizations['forecast'] = self.create_forecast_chart(sample_series, sample_forecast)
# Store visualization metadata
metadata = {
'analysis_type': analysis_type,
'timestamp': datetime.now().isoformat(),
'charts_generated': list(visualizations.keys()),
's3_bucket': self.s3_bucket
}
# Upload metadata
metadata_key = f"visualizations/metadata_{datetime.now().strftime('%Y%m%d_%H%M%S')}.json"
self.s3_client.put_object(
Bucket=self.s3_bucket,
Key=metadata_key,
Body=json.dumps(metadata, indent=2),
ContentType='application/json'
)
return visualizations
except Exception as e:
print(f"Error generating comprehensive visualizations: {e}")
return {}
def get_chart_url(self, chart_key: str) -> str:
"""Get public URL for a chart"""
try:
return f"https://{self.s3_bucket}.s3.amazonaws.com/{chart_key}"
except Exception as e:
print(f"Error generating chart URL: {e}")
return None
def list_available_charts(self) -> List[Dict]:
"""List all available charts in S3"""
try:
response = self.s3_client.list_objects_v2(
Bucket=self.s3_bucket,
Prefix='visualizations/'
)
charts = []
if 'Contents' in response:
for obj in response['Contents']:
if obj['Key'].endswith('.png'):
charts.append({
'key': obj['Key'],
'last_modified': obj['LastModified'],
'size': obj['Size'],
'url': self.get_chart_url(obj['Key'])
})
return sorted(charts, key=lambda x: x['last_modified'], reverse=True)
except Exception as e:
print(f"Error listing charts: {e}")
return []