Question

我制作了一个Python程序来遍历数据的csv文件并生成一个新的csv文件。为了减少运行时间，我尝试使用numba，下面是我的代码：

import csv
import timeit
from numba import jit
@jit(nopython=True)
def main():
    file = open("test_data.csv","r")
    newfile = open("rewrite5_gpu_output.csv",'w', newline='')
    writer = csv.writer(newfile)
    dataframe = []
    start = timeit.default_timer()
    for line in file:
        new_line = []
        line = line.split(",")
        for item in line:
            if item == '\n':
                pass
            elif item == "active":
                new_line.append(1) # 1 = active p53
            elif item == "inactive":
                new_line.append(0) # 0 = inactive p53
            else:
                try:
                    item = item.strip('\n')
                    item = float(item)
                    new_line.append(item)
                except:
                    pass
        dataframe.append(new_line)
        #print(new_line)
    writer.writerows(dataframe)
    newfile.close()
    stop = timeit.default_timer()
    print('Time: ', stop - start)

main()

但是，我得到了错误：

AssertionError: Failed in nopython mode pipeline (step: analyzing bytecode)
SETUP_EXCEPT(arg=32, lineno=23)

Answer 1

尝试注释掉“ try：”和“ except：”。在其他文章中，我读到了try-numba.jit不支持它可以在我的python脚本中工作。

br

AssertionError：在nopython模式管道中失败（步骤：分析字节码）

1 个答案: