Added comparison function as histogram operation (with colorized + info based string return)

README.md

@@ -125,8 +125,10 @@ histograms can be plotted via any compatible library, such as [mplhep][].
   * `*=`: Multiply by a scaler (not all storages) (`hist * scalar` and `scalar * hist` supported too)
   * `/=`: Divide by a scaler (not all storages) (`hist / scalar` supported too)
   * `.kind`: Either `bh.Kind.COUNT` or `bh.Kind.MEAN`, depending on storage
+  * `._storage_type`: Fetch histogram storage type class
   * `.sum(flow=False)`: The total count of all bins
   * `.project(ax1, ax2, ...)`: Project down to listed axis (numbers). Can also reorder axes.
+  * `.compare(second_hist)`: Compare the histogram with another histogram
   * `.to_numpy(flow=False, view=False)`: Convert to a NumPy style tuple (with or without under/overflow bins)
   * `.view(flow=False)`: Get a view on the bin contents (with or without under/overflow bins)
   * `.values(flow=False)`: Get a view on the values (counts or means, depending on storage)

src/boost_histogram/_internal/hist.py

@@ -337,6 +337,17 @@ def ndim(self) -> int:
         """
         return self._hist.rank()
 
+    def compare(self, hist2: "Histogram") -> str:
+        if not np.allclose(self.view().shape, hist2.view().shape):
+            return f"The histogram dimensions [{self.view().shape} and {hist2.view().shape}] are not equal."
+        if not np.allclose(self.view(), hist2.view()):
+            return f"The histogram contents :\n {self.view()} \nand\n {hist2.view()} \nare not equal."
+        if self._storage_type != hist2._storage_type:
+            return f"The storage types ({str(self._storage_type).rsplit('.', maxsplit=1)[-1][:-2]} and {str(hist2._storage_type).rsplit('.', maxsplit=1)[-1][:-2]}) are not equal."
+        if list(self.axes) != list(hist2.axes):
+            return f"The axes :\n {list(self.axes)} \nand\n {list(hist2.axes)} \nare not equal."
+        return ""
+
     def view(
         self, flow: bool = False
     ) -> Union["np.typing.NDArray[Any]", WeightedSumView, WeightedMeanView, MeanView]: