突然，Probit在PySAL (1.14.4)上不起作用

Question

直到上周，我还拥有一段代码，用于计算一系列事件发生的概率，但从本周开始，该代码不起作用。

代码如下：

import numpy as np
import pysal
import os

dbf = pysal.open('file.csv','r')

y = np.array([dbf.by_col('result')]).T

names_to_extract = ['dist', 'angle']
x = np.array([dbf.by_col(name) for name in names_to_extract]).T

id = np.array([dbf.by_col('incidenceId')]).T

model = pysal.spreg.probit.Probit(y, x, w=None, name_y='result', name_x=['dist','angle'], name_w=None, name_ds=None)

fin = np.column_stack((id, model.y, model.predy))

os.chdir("/destinyfolder")

np.savetxt('xg.csv', fin, delimiter=',', fmt='%d, %d, %f')

我得到了这个错误：

/usr/local/lib/python3.7/site-packages/pysal/__init__.py:65: VisibleDeprecationWarning: PySAL's API will be changed on 2018-12-31. The last release made with this API is version 1.14.4. A preview of the next API version is provided in the `pysalnext` package. The API changes and a guide on how to change imports is provided at https://migrating.pysal.org
  ), VisibleDeprecationWarning)
Traceback (most recent call last):
  File "xg.py", line 14, in <module>
    model = pysal.spreg.probit.Probit(y, x, w=None, name_y='result', name_x=['dist','angle'], name_w=None, name_ds=None)
  File "/usr/local/lib/python3.7/site-packages/pysal/spreg/probit.py", line 807, in __init__
    n = USER.check_arrays(y, x)
  File "/usr/local/lib/python3.7/site-packages/pysal/spreg/user_output.py", line 359, in check_arrays
    if not spu.spisfinite(i):
  File "/usr/local/lib/python3.7/site-packages/pysal/spreg/sputils.py", line 267, in spisfinite
    return np.isfinite(a.sum())
  File "/usr/local/lib/python3.7/site-packages/numpy/core/_methods.py", line 36, in _sum
    return umr_sum(a, axis, dtype, out, keepdims, initial)
TypeError: cannot perform reduce with flexible type

通常，我只得到DeprecationWarning，但代码可以工作。从今天开始，代码不能工作了。假设我当前使用的是pysal 1.14.4、numpy 1.16.2和scipy 1.2.1。我没有将我的代码更新到pysal 2.0，因为我不知道如何将此代码移植到新版本(这就是我首先获得DeprecationWarning的原因)。

这是文件：file.csv

你能帮我解决这个问题吗？

Answer 1

X中的条目是字符串，将它们转换为浮点数：

x = np.array(...).T.astype(np.float)

这首先会失败，因为file.csv中的"NULL“值，你需要过滤掉这些值，或者指定它们应该转换成什么浮点值。

在发布问题之前，先试着在谷歌上搜索一下，得到一个直接的答案。如果你在谷歌上搜索这个错误和"pysal 1.14.4"，你会立即找到这个解决方案：

TypeError: cannot perform reduce with flexible type