我有一个嵌套的for循环,它在4个时间窗口上执行10次迭代。伪代码是这样的: df=spark.read.parquet("very large dataset") time_window= 5 * i df_out=[operations performed in 10 different slices of the time window]
df_out.write.parquet("output path") 我知道每个循
下面是Python的代码片段。如stdout所示,活动线程不断增加,直到os抱怨线程太多。为什么在完成所有任务后,python不在thread_pool中恢复线程?from multiprocessing.dummy import Pool as ThreadPool passwhile True: data = list(range(1, 100))
thread_pool = ThreadPoo