为什么 numpy.histogram 尝试使用索引 -9223372036854775808（int64 的最小数量）获取值？

Question

我的任务是计算不同数字数组的直方图（我事先不知道）。我对相同双数的数组有麻烦，例如：

double_numbers = np.array([1e20] * 20)
np.histogram(double_numbers, bins=100)

我收到一个错误：

IndexError                                Traceback (most recent call last)
<ipython-input-37-51958daa5f2c> in <module>()
      1 double_numbers = np.array([1e20] * 20)
----> 2 np.histogram(double_numbers, bins=100)

/home/oleksandr/.pyenv/versions/2.7.8/envs/dr2.7.8/lib/python2.7/site-packages/numpy/lib/function_base.pyc in histogram(***failed resolving arguments***)
    795             # The index computation is not guaranteed to give exactly
    796             # consistent results within ~1 ULP of the bin edges.
--> 797             decrement = tmp_a_data < bin_edges[indices]
    798             indices[decrement] -= 1
    799             # The last bin includes the right edge. The other bins do not.

IndexError: index -9223372036854775808 is out of bounds for axis 1 with size 101

为什么 numpy.histogram 尝试使用索引 -9223372036854775808（int64 的最小值）获取值？ numpy有没有bug？

Answer 1

这是一个已知错误 https://github.com/numpy/numpy/issues/8627，已经存在大约 2 年了

为什么 numpy.histogram 尝试使用索引 -9223372036854775808（int64 的最小数量）获取值？

Why numpy.histogram try to get value with index -9223372036854775808 (min number of int64)?

python

numpy

histogram

index-error