make plots for truth information

.gitignore

 AutoDict*
 *.csv
+*.png
 .vscode/*
 *.root
 *.a
 *.o
 .venv/*
-*/__pycache__/*
\ No newline at end of file
+*/__pycache__/*

data/clusters.py

+import numpy as np
+import matplotlib.colors as colors
+import matplotlib.pyplot as plt
+
 from sklearn.datasets import make_blobs
 from tools import makeCSVFile
 
 class clusters:
-  def __init__(self, centers, std, n_samples, csv_path):
+  def __init__(self, centers, std, n_samples, bins, csv_path):
     self.centers = centers
     self.std = std
     self.n_samples = n_samples
     self.x, self.y, self.clusters = self.makeClusters()
+    self.bins = bins
     self.csv_path = csv_path
     self.dictionary()
-    
+    self.makePixelTruthPlot()
+    self.makeDetectorView()
 
   def makeClusters(self):
     nClusters = len(self.centers)
@@ -56,3 +62,72 @@ class clusters:
                    'simCluster_layer': [0 for _ in range(len(self.centers))]
                    }
      makeCSVFile(dictionary, self.csv_path)
+
+  def binnedClusters(self):
+    nclusters = len(self.clusters)
+    bins = self.bins
+    binned_clusters_content, binned_clusters_x, binned_clusters_y, binned_clusters_fraction = [], [], [], []
+
+    for clusters in range(0, nclusters):
+      content, xedges, yedges = np.histogram2d(self.clusters[clusters][0], self.clusters[clusters][1], bins=bins)
+      binned_clusters_content.append(content)
+      binned_clusters_x.append([0.5*(xedges[i] + xedges[i-1]) for i in range(1, len(xedges))])
+      binned_clusters_y.append([0.5*(yedges[i] + yedges[i-1]) for i in range(1, len(yedges))])
+
+    total_content, max_contribution = [], []
+    for x_bin in range(0, len(bins)-1):
+      total_content_x = []
+      for y_bin in range(0, len(bins)-1):
+        max_fraction = 0
+        max_fraction_cl = -1
+        total_bin_content = 0
+        for cluster in range(0, nclusters):
+          total_bin_content += binned_clusters_content[cluster][x_bin][y_bin]
+          if binned_clusters_content[cluster][x_bin][y_bin] > max_fraction:
+            max_fraction = binned_clusters_content[cluster][x_bin][y_bin]
+            max_fraction_cl = cluster
+        if total_bin_content == 0:
+          denominator = float("inf")
+        else:
+          denominator = total_bin_content
+        max_contribution.append([max_fraction_cl, max_fraction / denominator, total_bin_content])
+        total_content_x.append(denominator)
+      total_content.append(total_content_x)
+
+    for cluster in range(0, nclusters):
+      fractions = []
+      for x_bin in range(0, len(bins)-1):
+        x_fraction = []
+        for y_bin in range(0, len(bins)-1):
+          fraction = binned_clusters_content[cluster][x_bin][y_bin] / total_content[x_bin][y_bin]
+          x_fraction.append(fraction)
+        fractions.append(x_fraction)
+      binned_clusters_fraction.append(fractions)
+
+    return binned_clusters_content, binned_clusters_x, binned_clusters_y, max_contribution, binned_clusters_fraction
+  
+  def makePixelTruthPlot(self):
+    color = ['white', 'blue', 'green', 'red']
+    _, _, _, max_contribution, _ = self.binnedClusters()
+    size = len(self.bins) - 1
+    min_x = self.bins[0]
+    max_x = self.bins[-1]
+    min_y = min_x
+    max_y = max_x
+    cmap = colors.LinearSegmentedColormap.from_list("", color)
+    board = np.zeros((size, size))
+    fig, ax = plt.subplots(1,1)
+    for i in range(size):
+      for j in range(size):
+        board[i][j] = max_contribution[i + j * size][0]
+    ax.imshow(board, cmap=cmap, interpolation='nearest', extent=[min_x, max_x, min_y, max_y])
+    ax.invert_yaxis()
+    fig.savefig('pixelTruthPlot.png')
+    fig.clf()
+
+  def makeDetectorView(self):
+    plt.hist2d(self.x, self.y, bins=self.bins, norm=colors.LogNorm(), cmap='inferno')
+    plt.colorbar()
+    plt.xlim(self.bins[0], self.bins[-1])
+    plt.ylim(self.bins[0], self.bins[-1])
+    plt.savefig('detectorView.png')
\ No newline at end of file

data/config.yml

@@ -8,8 +8,8 @@ clusters:
 
 hits:
   bins:
-    start: -0.03
-    end: 0.03
+    start: -0.3
+    end: 0.3
     n_bins: 50
   csv_file: "csv/hits.csv"
 

data/makeSamples.py

@@ -16,9 +16,14 @@ if __name__ == '__main__':
     nSamples = config['clusters']['n_samples']
     centers = config['clusters']['centers']
 
+    bins = np.linspace(config['hits']['bins']['start'], 
+                       config['hits']['bins']['end'], 
+                       config['hits']['bins']['n_bins'])
+    
     x, y, clusters = clusters(config['clusters']['centers'],
                         config['clusters']['std'],
-                        config['clusters']['n_samples'], 
+                        config['clusters']['n_samples'],
+                        bins, 
                         config['clusters']['csv_file']).getHitsAndClusters()
 
     bins = np.linspace(config['hits']['bins']['start'], 

run-workflow.sh

@@ -32,7 +32,7 @@ parse_yml() {
 }
 
 print_step "Sourcing the LCG environment"
-source /cvmfs/sft.cern.ch/lcg/views/LCG_102/x86_64-centos8-gcc11-opt/setup.sh
+source /cvmfs/sft.cern.ch/lcg/views/LCG_105/x86_64-el9-gcc11-opt/setup.sh
 
 CURRENT_DIR=$(pwd)
 CSV_PATH=$(parse_yml ['csv_path'] | tr -d \', | awk '{print $2}')