把vic ASCII格式文本转为csv

代码记录：Change vic format file to standar file.  语言：python IDE;pycharm

"""""
==============================================================
get annual time series form vic format file
notice: format should be year month day prcp or single precp
2018.5.29 hjh ! Sep should be checked unix or pc? !
==============================================================
"""
import pandas as pd
import os
import time
start = time.time()

path = 'E:/Grid_pre_row/chirps/day/'  # file get in
out_file = 'F:/Syr/grid_pre_annual/chirps_result'  # outfile_name
out_file2 = 'G:/P/chirps/chirps.csv'  # standard outfile
start_year = 1981  # user input
end_year = 2016

files = [f for f in os.listdir(path) if f.startswith('data')]  # get a list of files
que = input('your input file format is standard? Y or N: ')
for file in files:  # read file one by one
    reader = pd.read_table(path+file, header=None, sep='\s+')
    if que == 'Y':
        reader.columns = ['year', 'month', 'day', 'prep']
    else:
        reader.columns = ['prep']
        x = pd.to_datetime(pd.date_range(str(start_year), str(end_year)+'-12-31'))
        reader.insert(0, 'ymd', x)
        reader['year'] = reader['ymd'].dt.year
        reader['month'] = reader['ymd'].dt.month
        reader['day'] = reader['ymd'].dt.day
        del reader['ymd']
        reader.reindex(columns=['year', 'month', 'day', 'prep'])
    reader_group = reader.groupby(['year', 'month'])['prep'].sum().reset_index()   # group file

    lat = float(file.split('_')[1])   # get file lat
    lon = float(file.split('_')[2])   # get file lon
    ll = [lat, lon]  # lat lon

    i = 0  # month slice start loc
    while True:
        month_value = reader_group.iloc[i:i+12, 2]   # 1-12 prcp, iloc is better
        year_index = reader_group.iloc[i:i+1, 0]     # strange.....
        reader_row = ll+list(year_index)+list(month_value)
        i = i+12
        with open(out_file, 'at') as f:
            print(*reader_row, sep=' ', file=f)  # *out put many objects
        if i >= (end_year-start_year+1)*12-1:  # last index in this file
            break
        print('reading'+'  '+str(file))
final_file = pd.read_table(out_file, header=None, sep='\s+')  # read result to Dataframe
column_names = ['lat', 'lon', 'year', 'one', 'two', 'three', 'four', 'five',
                'six', 'seven', 'eight', 'nine', 'ten', 'eleven', 'december']
final_file.columns = [column_names]  # give name to result
final_file = pd.DataFrame(final_file.apply(lambda x: pd.to_numeric(x, errors='ignore')))

'''A B C D represents (spring summer autumn winter)'''   # why must add values!

final_file['A'] = final_file['three'].values+final_file['four'].values+final_file['five'].values
final_file['B'] = final_file['six'].values+final_file['seven'].values+final_file['eight'].values
final_file['C'] = final_file['nine'].values+final_file['ten'].values+final_file['eleven'].values
final_file['D'] = final_file['one'].values+final_file['two'].values+final_file['december'].values
final_file['annual'] = final_file['A'].values + final_file['B'].values + final_file['C'].values + final_file['D'].values
final_file.to_csv(out_file2, index=None)
end = time.time()
a = end-start
print('programme run '+str(a)+' s')  # 600 S 51-07

把ascii格式的文本呢，转成一个csv文件，便于画图和分析数据等。