我正在使用在函数调用中使用一些十进制类型的args的库。但是我有numpy.int32类型变量。将其传递给函数调用时,出现以下错误:
TypeError:不支持从numpy.float32转换为Decimal
这意味着库尝试将我传递的参数转换为decimal.Decimal,但失败。请让我知道,在传递给库之前如何转换变量。谢谢
十进制接受本机Python类型。
numpyType.item()
返回相应的本机类型。
这是一个例子:
import decimal
import numpy as np
# other library function
# will raise exception if passed argument is not Decimal
def library(decimal_arg):
if type(decimal_arg) is not decimal.Decimal:
raise TypeError('Not a decimal.Decimal')
return decimal_arg
# wrapper with try block
def try_library(arg):
try:
print(library(arg))
except TypeError as e:
print(e)
def main():
# we have some numpy types
a = np.float32(1.1)
b = np.int32(2)
c = np.uint64(3000)
# passing the numpy type does not work
try_library(a)
# Decimal's constructor doesn't know conversion from numpy types
try:
try_library(decimal.Decimal(a))
except TypeError as e:
print(e)
# calling a.item() will return native Python type
# constructor of Decimal accepts a native type
# so it will work
try_library(decimal.Decimal(a.item()))
try_library(decimal.Decimal(b.item()))
try_library(decimal.Decimal(c.item()))
main()
# output:
# Not a decimal.Decimal
# Cannot convert 1.1 to Decimal
# 1.10000002384185791015625
# 2
# 3000
本文收集自互联网,转载请注明来源。
如有侵权,请联系 [email protected] 删除。
我来说两句