gallery_heatmap_correlation.cpp

This gallery example demonstrates creating an interactive correlation matrix heatmap using Plotly.cpp. It generates realistic business performance data across multiple metrics and displays their correlations using color coding and text annotations.

Features demonstrated:

• Symmetric correlation matrix generation with realistic business relationships
• Heatmap visualization with custom colorscale (RdBu for diverging data)
• Text annotations showing correlation coefficients on each cell
• Custom colorbar with centered zero point for proper correlation display
• Responsive layout with rotated axis labels and proper margins

The visualization helps identify relationships between different business metrics such as revenue, profit, marketing spend, and customer satisfaction.

Business Metrics Correlation Matrix

Date

2024

 
#include "plotly/plotly.hpp"
#include "utils/arg_parser.hpp"
#include <array>
#include <cmath>
#include <cstdio>
#include <random>
#include <string>
#include <vector>
 
auto main(int argc, char *argv[]) -> int {
  // Parse command line arguments
  auto args = parseGalleryArgs(argc, argv);
 
  plotly::Figure fig;
  fig.openBrowser(args.headless);
 
  // Define variable names
  std::vector<std::string> variables = {"Revenue",     "Profit",
                                        "Marketing",   "R&D",
                                        "Employees",   "Customer_Satisfaction",
                                        "Market_Share"};
  const size_t n = variables.size();
 
  // Generate realistic correlation matrix
  std::random_device rd;
  std::mt19937 gen(rd());
  std::uniform_real_distribution<double> correlationDist(-0.8, 0.9);
 
  std::vector<std::vector<double>> correlationMatrix(n, std::vector<double>(n));
 
  // Fill correlation matrix (symmetric)
  for (size_t i = 0; i < n; i++) {
    for (size_t j = 0; j < n; j++) {
      if (i == j) {
        correlationMatrix[i][j] = 1.0; // Perfect self-correlation
      } else if (i < j) {
        // Generate correlation with some business logic
        double corr = correlationDist(gen);
 
        // Add some realistic correlations
        if ((variables[i] == "Revenue" && variables[j] == "Profit") ||
            (variables[i] == "Marketing" && variables[j] == "Market_Share") ||
            (variables[i] == "R&D" &&
             variables[j] == "Customer_Satisfaction")) {
          corr = std::abs(corr) * 0.8 + 0.2; // Strong positive correlation
        }
 
        correlationMatrix[i][j] = corr;
        correlationMatrix[j][i] = corr; // Symmetric
      }
    }
  }
 
  // Create text annotations for correlation values
  std::vector<std::vector<std::string>> textMatrix(n,
                                                   std::vector<std::string>(n));
  for (size_t i = 0; i < n; i++) {
    for (size_t j = 0; j < n; j++) {
      std::array<char, 10> buffer;
      std::snprintf(buffer.data(), buffer.size(), "%.2f",
                    correlationMatrix[i][j]);
      textMatrix[i][j] = buffer.data();
    }
  }
 
  // Create heatmap trace
  plotly::Object trace = {
      {"type", "heatmap"},
      {"x", variables},
      {"y", variables},
      {"z", correlationMatrix},
      {"text", textMatrix},
      {"texttemplate", "%{text}"},
      {"textfont", {{"size", 12}, {"color", "white"}}},
      {"colorscale", "RdBu"},
      {"zmid", 0.0},
      {"showscale", true},
      {"colorbar",
       {{"title", "Correlation Coefficient"}, {"titleside", "right"}}}};
 
  // Create layout
  plotly::Object layout = {
      {"title",
       {{"text", "Business Metrics Correlation Matrix"},
        {"font", {{"size", 16}}}}},
      {"xaxis",
       {{"title", "Variables"}, {"side", "bottom"}, {"tickangle", 45}}},
      {"yaxis",
       {
           {"title", "Variables"}, {"autorange", "reversed"}
           // Reverse y-axis for matrix display
       }},
      {"width", 800},
      {"height", 700},
      {"margin", {{"l", 150}, {"r", 100}, {"t", 100}, {"b", 150}}}};
 
  // Create the plot
  std::vector<plotly::Object> data = {trace};
  fig.newPlot(data, layout);
 
  if (!args.headless) {
    fig.waitClose();
  } else {
    // Save image instead of opening browser
    plotly::Object imageOpts = {{"format", "png"},
                                {"width", 800},
                                {"height", 700},
                                {"filename", "heatmap_correlation"}};
    fig.downloadImage(imageOpts);
  }
 
  return 0;
}