[Python] Pandas df.iterrow 平行处理

#import pakageimport pandas as pdimport multiprocessing#指定主机有多少CPU，并把df 平均分配sizenum_processes = multiprocessing.cpu_count()chunk_size = int(df.shape[0]/num_processes)#将df 拆解成 num_processes 个 chunk 的df listchunks = [df.ix[df.index[i:i + chunk_size]] for i in range(0, df.shape[0], chunk_size)]

#定义 functiondef myfunc(d):   return d * dpool = multiprocessing.Pool(processes=num_processes)result = pool.map(myfunc, chunks)