4 from numbers import Number
5 from typing import Generic, Iterable, List, Optional, Tuple, TypeVar
7 T = TypeVar("T", bound=Number)
10 class SimpleHistogram(Generic[T]):
12 # Useful in defining wide open bottom/top bucket bounds:
13 POSITIVE_INFINITY = math.inf
14 NEGATIVE_INFINITY = -math.inf
16 def __init__(self, buckets: List[Tuple[T, T]]):
17 from math_utils import RunningMedian
20 for start_end in buckets:
21 if self._get_bucket(start_end[0]) is not None:
22 raise Exception("Buckets overlap?!")
23 self.buckets[start_end] = 0
25 self.median = RunningMedian()
31 def n_evenly_spaced_buckets(
35 ) -> List[Tuple[T, T]]:
37 stride = int((max_bound - min_bound) / n)
39 raise Exception("Min must be < Max")
40 for bucket_start in range(min_bound, max_bound, stride):
41 ret.append((bucket_start, bucket_start + stride))
44 def _get_bucket(self, item: T) -> Optional[Tuple[T, T]]:
45 for start_end in self.buckets:
46 if start_end[0] <= item < start_end[1]:
50 def add_item(self, item: T) -> bool:
51 bucket = self._get_bucket(item)
55 self.buckets[bucket] += 1
57 self.median.add_number(item)
58 if self.maximum is None or item > self.maximum:
60 if self.minimum is None or item < self.minimum:
64 def add_items(self, lst: Iterable[T]) -> bool:
67 all_true = all_true and self.add_item(item)
70 def __repr__(self, label_formatter='%10s') -> str:
71 from text_utils import bar_graph
73 max_population: Optional[int] = None
74 for bucket in self.buckets:
75 pop = self.buckets[bucket]
77 last_bucket_start = bucket[0] # beginning of range
78 if max_population is None or pop > max_population:
79 max_population = pop # bucket with max items
82 if max_population is None:
85 for bucket in sorted(self.buckets, key=lambda x: x[0]):
86 pop = self.buckets[bucket]
90 (pop / max_population),
96 label = f'{label_formatter}..{label_formatter}' % (start, end)
97 txt += f'{label:20}: ' + bar + f"({pop/self.count*100.0:5.2f}% n={pop})\n"
98 if start == last_bucket_start: