| 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748 |
- import multiprocessing
- import os.path
- from utils.file.trans_methods import *
- from utils.systeminfo.sysinfo import get_available_cpu_count_with_percent
- class WaveTrans(object):
- def __init__(self, field_code, read_path, save_path: str):
- self.field_code = field_code
- self.read_path = read_path
- self.save_path = save_path
- def get_data(self, file_path):
- df = pd.read_csv(file_path, encoding=detect_file_encoding(file_path), header=None)
- data = [i for i in df[0].values]
- filename = os.path.basename(file_path)
- wind_num = filename.split('_')[1]
- cedian = '齿轮箱' + filename.split('_齿轮箱')[1].split('_Time')[0]
- cedian_time = filename.split('风机_')[1].split('_齿轮箱')[0].replace('_', ':')
- name_tmp = 'Time_' + filename.split('Time_')[1].split('_cms')[0]
- pinlv = name_tmp[0:name_tmp.rfind('_')]
- zhuansu = name_tmp[name_tmp.rfind('_') + 1:]
- df = pd.DataFrame()
- df['风机编号'] = [wind_num, wind_num]
- df['时间'] = [cedian_time, cedian_time]
- df['频率'] = [pinlv, pinlv]
- df['测点'] = ['转速', cedian]
- df['数据'] = [[float(zhuansu)], data]
- return df
- def run(self):
- all_files = read_files(self.read_path, ['csv'])
- # 最大取系统cpu的 1/2
- split_count = get_available_cpu_count_with_percent(1 / 2)
- with multiprocessing.Pool(split_count) as pool:
- dfs = pool.starmap(self.get_data, [(i,) for i in all_files])
- df = pd.concat(dfs, ignore_index=True, copy=False)
- df.drop_duplicates(subset=['风机编号', '时间', '频率', '测点'], keep='last')
- df.to_csv(os.path.join(self.save_path, self.field_code + '.csv'), index=False, encoding='utf8')
|
