我正在尝试通过将DF类型转换为更高效的类型来优化我们的内存使用（今天，float64占了很多，它消耗8位并且非常低效）。
我写了下面的代码（如果您认为我认为这会从乞讨开始起作用的话是错误的，请告诉我）
@staticmethod
def optimize_float(series):
low_consumption = series.astype('float16')
if any(low_consumption.isin([np.inf])):
medium_consumption = series.astype('float32')
if any(medium_consumption.isin([np.inf])):
high_consumption = series.astype('float64')
return high_consumption
return medium_consumption
return low_consumption
此代码实际上大大减少了DF的内存使用量。
但是稍后在代码中，我得到以下错误：
Traceback (most recent call last):
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask/app.py", line 1985, in wsgi_app
response = self.handle_exception(e)
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask_restful/__init__.py", line 273, in error_router
return original_handler(e)
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask_cors/extension.py", line 161, in wrapped_function
return cors_after_request(app.make_response(f(*args, **kwargs)))
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask/app.py", line 1540, in handle_exception
reraise(exc_type, exc_value, tb)
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask/_compat.py", line 32, in reraise
raise value.with_traceback(tb)
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask/app.py", line 1982, in wsgi_app
response = self.full_dispatch_request()
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask/app.py", line 1614, in full_dispatch_request
rv = self.handle_user_exception(e)
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask_restful/__init__.py", line 273, in error_router
return original_handler(e)
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask_cors/extension.py", line 161, in wrapped_function
return cors_after_request(app.make_response(f(*args, **kwargs)))
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask/app.py", line 1518, in handle_user_exception
return handler(e)
File "/Users//Desktop/code/romee/autoai/server/__init__.py", line 186, in handle_general_exeption
raise error
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask_restful/__init__.py", line 270, in error_router
return self.handle_error(e)
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask/app.py", line 1612, in full_dispatch_request
rv = self.dispatch_request()
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask_debugtoolbar/__init__.py", line 125, in dispatch_request
return view_func(**req.view_args)
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask_restful/__init__.py", line 484, in wrapper
return self.make_response(data, code, headers=headers)
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask_restful/__init__.py", line 513, in make_response
resp = self.representations[mediatype](data, *args, **kwargs)
File "/Users//.local/share/virtualenvs/romee-dkO78zKu/lib/python3.6/site-packages/flask_restful/representations/json.py", line 21, in output_json
dumped = dumps(data, **settings) + "\n"
File "/Users//.pyenv/versions/3.6.5/lib/python3.6/json/__init__.py", line 238, in dumps
**kw).encode(obj)
File "/Users//.pyenv/versions/3.6.5/lib/python3.6/json/encoder.py", line 201, in encode
chunks = list(chunks)
File "/Users//.pyenv/versions/3.6.5/lib/python3.6/json/encoder.py", line 430, in _iterencode
yield from _iterencode_dict(o, _current_indent_level)
File "/Users//.pyenv/versions/3.6.5/lib/python3.6/json/encoder.py", line 404, in _iterencode_dict
yield from chunks
File "/Users//.pyenv/versions/3.6.5/lib/python3.6/json/encoder.py", line 325, in _iterencode_list
yield from chunks
File "/Users//.pyenv/versions/3.6.5/lib/python3.6/json/encoder.py", line 404, in _iterencode_dict
yield from chunks
File "/Users//.pyenv/versions/3.6.5/lib/python3.6/json/encoder.py", line 325, in _iterencode_list
yield from chunks
File "/Users//.pyenv/versions/3.6.5/lib/python3.6/json/encoder.py", line 404, in _iterencode_dict
yield from chunks
File "/Users//.pyenv/versions/3.6.5/lib/python3.6/json/encoder.py", line 396, in _iterencode_dict
yield _floatstr(value)
File "/Users//.pyenv/versions/3.6.5/lib/python3.6/json/encoder.py", line 241, in floatstr
repr(o))
ValueError: Out of range float values are not JSON compliant: -inf
stacktrace不是很有用-因为它没有指出此错误来自 my 代码中的位置。
我测试了optimize float方法，看它没有返回np.inf值（if any(medium_consumption.isin([np.inf]))应该会解决这个问题。
我想念什么？
编辑：
我注意到在尝试向客户发送本系列中describe函数的内容时会发生这种情况。
可复制的示例：
df = pd.DataFrame({'a':[11111,22222,3333]})
df['a'].describe()
Out[84]:
count 3.000000
mean 12222.000000
std 9493.383011
min 3333.000000
25% 7222.000000
50% 11111.000000
75% 16666.500000
max 22222.000000
Name: a, dtype: float64
df['a'].astype('float16').describe()
Out[85]:
count 3.000000e+00
mean 1.222400e+04
std inf
min 3.332000e+03
25% 7.222000e+03
50% 1.111200e+04
75% 1.666800e+04
max 2.222400e+04
Name: a, dtype: float64
现在的问题是，我该如何舍入从describe返回的值，所以它永远不会是inf（但仍保留列float16）