python - 重命名 pcraster Mapstack

Question

我有一个文件夹，里面装满了 20 年降水 pcraster map 堆栈，我已经设法从原始 netcdf 文件中提取出我感兴趣区域的降水值，并将其重命名为该值以避免混淆

precip.19810101 
precip.19810102 
precip.19810103 
precip.19810104 
precip.19810105
...
precip.20111231

但在那之后，我想根据这个日期序列将所有文件重命名为 pcraster mapstack

precip00.001
precip00.002
precip00.003
precip00.004
...

我是Python初学者，有什么帮助或例子可以帮助我弄清楚如何做到这一点吗？ 谢谢

Answer 1

这是我根据我曾经编写的一些旧 Python 脚本整理的内容:

#! /usr/bin/env python
# Rename PCRaster map stack with names following prefix.yyymmmdd to stack with valid
# PCRaster time step numbers
# Johan van der Knijff
#
# Example input stack:
#
# precip.19810101
# precip.19810102
# precip.19810103
# precip.19810104
# precip.19810105
#
# Then run script with following arguments:
#
# python renpcrstack.py precip 1
#
# Result:
#
# precip00.001
# precip00.002
# precip00.003
# precip00.004
# precip00.005
#

import sys
import os
import argparse
import math
import datetime
import glob

# Create argument parser
parser = argparse.ArgumentParser(
    description="Rename map stack")

def parseCommandLine():
    # Add arguments
    parser.add_argument('prefix',
                        action="store",
                        type=str,
                        help="prefix of input map stack (also used as output prefix)")
    parser.add_argument('stepStartOut',
                        action="store",
                        type=int,
                        help="time step number that is assigned to first map in output stack")

    # Parse arguments
    args = parser.parse_args()

    return(args)

def dateToJulianDay(date):

    # Calculate Julian Day from date
    # Source: https://en.wikipedia.org/wiki/Julian_day#Converting_Julian_or_Gregorian_calendar_date_to_Julian_day_number

    a = (14 - date.month)/12
    y = date.year + 4800 - a
    m = date.month +12*a - 3

    JulianDay = date.day + math.floor((153*m + 2)/5) + 365*y + math.floor(y/4) \
        - math.floor(y/100) + math.floor(y/400) - 32045

    return(JulianDay)

def genStackNames(prefix,start,end, stepSize):
    # Generate list with names of all maps
    # map name is made up of 11 characters, and chars 8 and 9 are
    # separated by a dot. Name starts with prefix, ends with time step
    # number and all character positions in between are filled with zeroes

    # define list that will contain map names
    listMaps = []

    # Count no chars prefix
    charsPrefix = len(prefix)

    # Maximum no chars needed for suffix (end step)
    maxCharsSuffix = len(str(end))

    # No of free positions between pre- and suffix
    noFreePositions = 11 - charsPrefix - maxCharsSuffix

    # Trim prefix if not enough character positions are available 
    if noFreePositions < 0:
        # No of chars to cut from prefix if 11-char limit is exceeded
        charsToCut = charsPrefix + maxCharsSuffix - 11
        charsToKeep = charsPrefix - charsToCut

        # Updated prefix
        prefix = prefix[0:charsToKeep]

        # Updated prefix length
        charsPrefix = len(prefix)

    # Generate name for each step

    for i in range(start,end + 1,stepSize):

        # No of chars in suffix for this step
        charsSuffix = len(str(i))

        # No of zeroes to fill
        noZeroes = 11 - charsPrefix - charsSuffix

        # Total no of chars right of prefix
        charsAfterPrefix = noZeroes + charsSuffix

        # Name of map

        thisName = prefix + (str(i)).zfill(charsAfterPrefix)
        thisFile = thisName[0:8]+"." + thisName[8:11]

        listMaps.append(thisFile)

    return listMaps    

def main():
    # Parse command line arguments
    args = parseCommandLine()
    prefix = args.prefix
    stepStartOut = args.stepStartOut

    # Glob pattern for input maps: prefix + dot + 8 char extension
    pattern = prefix + ".????????"

    # Get list of all input maps based on glob pattern
    mapsIn = glob.glob(pattern)

    # Set time format
    tfmt = "%Y%m%d"

    # Set up dictionary that will act as lookup table between Julian Days (key) 
    # and Date string
    jDayDate = {}

    for map in mapsIn:
        baseNameIn = os.path.splitext(map)[0]
        dateIn = os.path.splitext(map)[1].strip(".")

        # Convert to date / time format
        dt = datetime.datetime.strptime(dateIn, tfmt)

        # Convert date to Julian day number
        jDay = int(dateToJulianDay(dt))

        # Store as key-value pair in dictionary
        jDayDate[jDay] = dateIn

    # Number of input maps (equals number of key-value pairs)
    noMaps = len(jDayDate)

    # Create list of names for output files 
    mapNamesOut = genStackNames(prefix, stepStartOut, noMaps + stepStartOut -1, 1)

    # Iterate over Julian Days (ascending order)

    i = 0

    for key in sorted(jDayDate):
        # Name of input file
        fileIn = prefix + "."+ jDayDate[key]

        # Name of output file
        fileOut = mapNamesOut[i]

        # Rename file
        os.rename(fileIn, fileOut)

        print("Renamed " + fileIn + " ---> " + fileOut)

        i += 1

main()

(或者从 my Github Gist 下载代码。)

您可以从命令行运行它，使用 map 堆栈的前缀和第一个输出 map 的编号作为参数，例如:

python renpcrmaps.py precip 1

请注意，脚本会重命名文件，因此请确保复制原始 map 堆栈，以防出现问题(我对此仅进行了一些非常有限的测试!) .

此外，该脚本假定一个非稀疏输入 map 堆栈，即在每日 map 的情况下，每一天都存在一个输入 map 。如果缺少天数，输出 map 的编号将不是您所期望的。

所有日期到儒略日的内部转换在这里可能有点矫枉过正，但是一旦您开始进行更高级的转换，它确实会让事情变得更容易，因为它为您提供比日期字符串更容易操作的十进制数字。