add some utilities for reading/plotting

SIRF_data_preparation/evaluation_utilities.py

@@ -0,0 +1,38 @@
+"""
+Some utilities for plotting objectives and metrics
+"""
+import csv
+import os
+import numpy
+from typing import Iterator
+import matplotlib.pyplot as plt
+from petric import QualityMetrics
+import sirf.STIR as STIR
+
+def read_objectives(datadir='.') -> numpy.array:
+    """Reads objectives.csv and returns as 2d array"""
+    with open(os.path.join(datadir, 'objectives.csv'), newline='') as csvfile:
+        reader = csv.reader(csvfile)
+        next(reader) # skip first line
+        objs = numpy.array([(float(row[0]), float(row[1])) for row in reader])
+        return objs
+
+
+def get_metrics(qm: QualityMetrics, iters: Iterator, srcdir='.') -> numpy.array :
+    """Read 'iter*.hv' images from datadir, compute metrics and return as 2d array"""
+    m = []
+    for iter in iters:
+        im = STIR.ImageData(os.path.join(srcdir,f"iter_{iter:04d}.hv"))
+        m.append([*qm.evaluate(im).values()])
+    return numpy.array(m)
+#%%
+def plot_metrics(iters: Iterator, m: numpy.array, labels=[], suffix=""):
+    """Make 2 subplots of metrics"""
+    ax = plt.subplot(121)
+    plt.plot(iters, m[:,0], label=labels[0]+suffix)
+    plt.plot(iters, m[:,1], label=labels[1]+suffix)
+    ax.legend()
+    ax = plt.subplot(122)
+    for i in range(2, m.shape[1]):
+        plt.plot(iters, m[:,i], label=labels[i]+suffix)
+    ax.legend()

petric.py

@@ -120,7 +120,7 @@ def evaluate(self, test_im: STIR.ImageData) -> dict[str, float]:
             for voi_name, voi_indices in sorted(self.voi_indices.items())}
         return {**whole, **local}
 
-    def keys():
+    def keys(self):
         l = ["RMSE_whole_object", "RMSE_background"]
         for voi_name in sorted(self.voi_indices.keys()):
             l.append(f"AEM_VOI_{voi_name}")