thunlp / THULAC-Python

An Efficient Lexical Analyzer for Chinese
MIT License
2.02k stars 336 forks source link

module 'time' has no attribute 'clock' #100

Open BrandNewJimZhang opened 4 years ago

BrandNewJimZhang commented 4 years ago

在Python 3.8中,time.clock()已经被移除了。但是切割句子时仍然使用了这个。

dovecho commented 4 years ago

Simply change start = time.clock() to start = time.perf_counter() in the file: thulac/character/CBTaggingDecoder.py, the program will run well

HannibalXZX commented 4 years ago

我也有这个问题,但是在2.7版本可以运行。你可以尝试改成2.7,也可以把相应的函数替换为3.8版本的

use time.perf_counter or time.process_time instead time.clock()

改完以后,我这边3.8版本,运行成功。

fncokg commented 3 years ago

Simply change start = time.clock() to start = time.perf_counter() in the file: thulac/character/CBTaggingDecoder.py, the program will run well

Modifying the source code isn't a good practice, since that means you cannot migrate your code to a another device. Try this instead(before the official fix this):

import time
if not hasattr(time, 'clock'):
    setattr(time,'clock',time.perf_counter)

Then your code will run well no matter on a Python<3.8 or a Python>=3.8.

YunfangHou commented 2 years ago

Simply change start = time.clock() to start = time.perf_counter() in the file: thulac/character/CBTaggingDecoder.py, the program will run well

Modifying the source code isn't a good practice, since that means you cannot migrate your code to a another device. Try this instead(before the official fix this):

import time
if not hasattr(time, 'clock'):
    setattr(time,'clock',time.perf_counter)

Then your code will run well no matter on a Python<3.8 or a Python>=3.8.

It works. Thank you.