More mypy cleanup... ugh.
[python_utils.git] / histogram.py
index 3391b0b3c51d33164b6d173d59a7b4a1b3c9788b..87ef77340fd036b8ea29d5ddf3def5d902db3688 100644 (file)
@@ -15,6 +15,7 @@ class SimpleHistogram(Generic[T]):
 
     def __init__(self, buckets: List[Tuple[T, T]]):
         from math_utils import RunningMedian
+
         self.buckets = {}
         for start_end in buckets:
             if self._get_bucket(start_end[0]) is not None:
@@ -28,9 +29,9 @@ class SimpleHistogram(Generic[T]):
 
     @staticmethod
     def n_evenly_spaced_buckets(
-            min_bound: T,
-            max_bound: T,
-            n: int,
+        min_bound: T,
+        max_bound: T,
+        n: int,
     ) -> List[Tuple[T, T]]:
         ret = []
         stride = int((max_bound - min_bound) / n)
@@ -66,8 +67,7 @@ class SimpleHistogram(Generic[T]):
             all_true = all_true and self.add_item(item)
         return all_true
 
-    def __repr__(self,
-                 label_formatter='%10s') -> str:
+    def __repr__(self, *, width: int = 80, label_formatter: str = None) -> str:
         from text_utils import bar_graph
 
         max_population: Optional[int] = None
@@ -82,18 +82,56 @@ class SimpleHistogram(Generic[T]):
         if max_population is None:
             return txt
 
-        for bucket in sorted(self.buckets, key=lambda x : x[0]):
-            pop = self.buckets[bucket]
+        max_label_width = None
+        lowest_start = None
+        highest_end = None
+        for bucket in sorted(self.buckets, key=lambda x: x[0]):
+            start = bucket[0]
+            if lowest_start is None:
+                lowest_start = start
+            end = bucket[1]
+            if highest_end is None or end > highest_end:
+                highest_end = end
+            if label_formatter is None:
+                if type(start) == int and type(end) == int:
+                    label_formatter = '%d'
+                elif type(start) == float and type(end) == float:
+                    label_formatter = '%.2f'
+                else:
+                    label_formatter = '%s'
+            label = f'[{label_formatter}..{label_formatter}): ' % (start, end)
+            label_width = len(label)
+            if max_label_width is None or label_width > max_label_width:
+                max_label_width = label_width
+            if start == last_bucket_start:
+                break
+        sigma_label = f'[{label_formatter}..{label_formatter}): ' % (
+            lowest_start,
+            highest_end,
+        )
+        if len(sigma_label) > max_label_width:
+            max_label_width = len(sigma_label)
+        bar_width = width - (max_label_width + 16)
+
+        for bucket in sorted(self.buckets, key=lambda x: x[0]):
             start = bucket[0]
             end = bucket[1]
+            label = f'[{label_formatter}..{label_formatter}): ' % (start, end)
+            pop = self.buckets[bucket]
             bar = bar_graph(
                 (pop / max_population),
-                include_text = False,
-                width = 58,
-                left_end = "",
-                right_end = "")
-            label = f'{label_formatter}..{label_formatter}' % (start, end)
-            txt += f'{label:20}: ' + bar + f"({pop/self.count*100.0:5.2f}% n={pop})\n"
+                include_text=False,
+                width=bar_width,
+                left_end="",
+                right_end="",
+            )
+            txt += label.rjust(max_label_width)
+            txt += bar
+            txt += f"({pop/self.count*100.0:5.2f}% n={pop})\n"
             if start == last_bucket_start:
                 break
+        txt += '-' * width + '\n'
+        txt += sigma_label.rjust(max_label_width)
+        txt += ' ' * (bar_width - 2)
+        txt += f'Σ=(100.00% n={self.count})\n'
         return txt