处理adcp

import pandas as pd
import numpy as np
import datetime as dt
import glob
import os
import re

from itertools import islice
from matplotlib import pyplot as plt


#
# #打开文件
def get_wenjian(file, n):
    return [x.strip() for x in islice(file, n)]


def chuli_head(line):
    res = {}
    res['ens'] = int(line[0].split()[7])
    res['date'] = dt.datetime(year=2000 + int(line[0].split()[0]), month=int(line[0].split()[1]),
                              day=int(line[0].split()[2]),
                              hour=int(line[0].split()[3]), minute=int(line[0].split()[4]),
                              second=int(line[0].split()[5]))
    # print(res['date'])
    # res['h'] = round(float(line[1].split()[8]), 3)
    s1 = round(float(line[1].split()[8]), 3)
    s2 = round(float(line[1].split()[9]), 3)
    s3 = round(float(line[1].split()[10]), 3)
    s4 = round(float(line[1].split()[11]), 3)
    s = np.array([s1, s2, s3, s4])
    d1 = np.mean(s)
    res['h'] =d1
    # s = np.array(line[1].split()[8], line[1].split()[9], line[1].split()[10], line[1].split()[11])
    # res['h'] = np.mean(s)  # 平均深度
    res['b'] = round(float(line[2].split()[0]), 2)
    res['lat'] = float(line[3].split()[0]) if float(line[3].split()[0]) != 30000. else 'NA'
    res['lon'] = float(line[3].split()[1]) if float(line[3].split()[1]) != 30000. else 'NA'
    res['top_q'] = round(float(line[4].split()[1]), 3)  # 顶层流量
    res['b_q'] = round(float(line[4].split()[2]), 3)  # 底部流量
    res['nbins'] = int(line[5].split()[0])  # 层数

    return res


def chuli_shuju(ens, ensnum, ensdist, ensh, enslat, enslon, topq, botq,date):
    res = {}
    df = pd.DataFrame([x.split() for x in ens], columns=['hb', 'v', 'd', 'v1', 'v2', 'v3', 'v4',
                                                         'bs1', 'bs2', 'bs3', 'bs4', 'percgood', 'q'], dtype='float')
    rmcols = ['v1', 'v2', 'v3', 'v4', 'percgood', 'q']
    df.drop(rmcols, inplace=True, axis=1)  # 删去不需要用到的数据
    df = df.replace(dict.fromkeys([-32768, 2147483647, 255], np.nan))
    df = df.dropna()
    df['bs'] = df[['bs1', 'bs2', 'bs3', 'bs4']].mean(axis=1)  # 平均4探头回波强度
    df.drop(['bs1', 'bs2', 'bs3', 'bs4'], inplace=True, axis=1)  # 删去4探头回波强度
    df['ens'] = ensnum  # 剖面号
    df['date'] = date  # 添加日期
    df['dist'] = ensdist
    df['h'] = ensh  # 使用头部信息中的水深
    df['lat'] = enslat  # 头部信息纬度信息
    df['lon'] = enslon  # 头部信息经度信息
    df['v'] = df['v'] * 0.01
    to_round = [col for col in df.columns if not 'lat' in col and not 'lon' in col]
    # print(to_round)
    df[to_round] = df[to_round].round(3)  # 保留3位小数
    res = pd.melt(df, id_vars=['ens', 'dist', 'lon', 'lat', 'hb', 'h','date'])  # 合并出现 value 和variable 即将 v,d,bs 合并成一列 #
    # 出现的variable为列名，value为分别的值，id_vars 为不合并列
    # print(res)
    return res


def file_proc(path_in, path_out):  # 主函数
    with open(path_in, "r") as f:
        f.readline()  # 去除前三行无效信息
        f.readline()
        f.readline()
        opr = {}
        df = pd.DataFrame()  # 建立数据表
        head = get_wenjian(f, 6)  # 获取文件头6行信息并释放迭代器
        while head:
            opr = chuli_head(head)
            chunk = get_wenjian(f, opr['nbins'])  # 获得前6行之后的层数行文件信息
            ens = chuli_shuju(chunk, opr['ens'], opr['b'], opr['h'], opr['lat'], opr['lon'], opr['top_q'], opr['b_q'],opr['date'])
            df = df.append(ens, ignore_index=True)
            head = get_wenjian(f, 6)  # 进去下一个文件信息循环，即层数 + 头部信息
        # print(chunk)
        # print(ens)
        # print(df)
        df.to_csv(path_out, index=False, na_rep='-32768')
        bsf = df.loc[df['variable'] == 'bs', ['dist', 'hb', 'value', 'h']]

        bsf['hb'] = bsf['hb'].astype(float)
        bsf['h'] = bsf['h'].astype(float)
        bsf['value'] = bsf['value'].astype(float)

        print('Finished ' + path_in)
    return


if __name__ == "__main__":
    ff = glob.glob("*_ASC.txt")
    if not ff:
        print('No files to convert.')
        exit()
    else:
        print("Detected ASCII *_ASC.txt files: 
", "
".join(ff))
        for f in ff:
            file_in = f
            file_out = re.sub(r'(?i)txt', 'csv', f)
            file_proc(file_in, file_out)