gpt4 book ai didi

python - 传递超过 23 个输入文件时 gdal_calc amin 失败

转载 作者:太空狗 更新时间:2023-10-29 20:29:58 25 4
gpt4 key购买 nike

我编写了一个调用 gdal_calc.py 的 R 函数来计算 RasterStack(一系列输入栅格文件)中的像素最小值。我这样做是因为它比大型栅格的 raster::min 快得多。该函数适用于最多 23 个文件,但在传递 24 个或更多文件时会引发警告,并返回一个用零填充的输出栅格。

由于 R 只是准备对 python gdal_calc.py 进行系统调用,所以这个问题并不特定于 R,我鼓励 python/numpy 爱好者继续阅读。

这是函数。最终 gdal_calc 调用的结构显示在本文底部的有问题的用法引发的警告消息中。

gdal_min <- function(infile, outfile, return_raster=FALSE, overwrite=FALSE) {
require(rgdal)
if(return_raster) require(raster)
# infile: The multiband raster file (or a vector of paths to multiple
# raster files) for which to calculate cell minima.
# outfile: Path to raster output file.
# return_raster: (logical) Should the output raster be read back into R?
# overwrite: (logical) Should outfile be overwritten if it exists?
gdal_calc <- Sys.which('gdal_calc.py')
if(gdal_calc=='') stop('gdal_calc.py not found on system.')
if(file.exists(outfile) & !overwrite)
stop("'outfile' already exists. Use overwrite=TRUE to overwrite.",
call.=FALSE)
nbands <- sapply(infile, function(x) nrow(attr(GDALinfo(x), 'df')))
if(length(infile) > 26 || nbands > 26) stop('Maximum number of inputs is 26.')
if(length(nbands) > 1 & any(nbands > 1))
warning('One or more rasters have multiple bands. First band used.')

if(length(infile)==1) {
inputs <- paste0('-', LETTERS[seq_len(nbands)], ' ', infile, ' --',
LETTERS[seq_len(nbands)], '_band ', seq_len(nbands), collapse=' ')
n <- nbands
} else {
inputs <- paste0('-', LETTERS[seq_along(nbands)], ' ', infile, ' --',
LETTERS[seq_along(nbands)], '_band 1', collapse=' ')
n <- length(infile)
}

message('Calculating minimum and writing to ', basename(outfile))
system2('python',
args=c(gdal_calc, inputs,
sprintf("--outfile=%s", outfile),
sprintf('--calc="amin([%s], axis=0)"',
paste0(LETTERS[seq_len(n)], collapse=',')),
'--co="COMPRESS=LZW"',
if(overwrite) '--overwrite'),
stdout=FALSE
)
if(return_raster) raster(outfile) else invisible(NULL)
}

这是一个虚拟栅格网格,写入当前工作目录中的 'test.tif'。为简单起见,并为了证明问题不是由特殊的单个栅格引起的,我将同一个输入文件多次传递给 gdal_calc,而不是传递多个不同的文件。

library(raster)
writeRaster(raster(matrix(runif(100), 10)), 'test.tif')

该函数适用于 23 个输入文件:

r <- gdal_min(rep('test.tif', 23), f_out <- tempfile(fileext='.tif'),
return_raster=TRUE)
plot(r)

...但不适用于 ≥ 24:

r <- gdal_min(rep('test.tif', 24), f_out <- tempfile(fileext='.tif'),
return_raster=TRUE)
## Calculating minimum and writing to file2310254bcc.tif
## Warning message:
## running command '"python" C:\Python33\Scripts\GDAL_C~1.PY -A test.tif --A_band 1 -B test.tif --B_band 1 -C test.tif --C_band 1 -D test.tif --D_band 1 -E test.tif --E_band 1 -F test.tif --F_band 1 -G test.tif --G_band 1 -H test.tif --H_band 1 -I test.tif --I_band 1 -J test.tif --J_band 1 -K test.tif --K_band 1 -L test.tif --L_band 1 -M test.tif --M_band 1 -N test.tif --N_band 1 -O test.tif --O_band 1 -P test.tif --P_band 1 -Q test.tif --Q_band 1 -R test.tif --R_band 1 -S test.tif --S_band 1 -T test.tif --T_band 1 -U test.tif --U_band 1 -V test.tif --V_band 1 -W test.tif --W_band 1 -X test.tif --X_band 1 --outfile=C:\Users\John\AppData\Local\Temp\RtmpK4heMM\file2310254bcc.tif --calc="amin([A,B,C,D,E,F,G,H,I,J,K,L,M,N,O,P,Q,R,S,T,U,V,W,X], axis=0)" --co="COMPRESS=LZW"' had status 1

直接从终端执行后者时给出的回溯是:

"python" C:\Python33\Scripts\GDAL_C~1.PY -A test.tif --A_band 1 -B test.tif --B_band 1 -C test.tif --C_band 1 -D test.tif --D_band 1 -E test.tif --E_band 1 -F test.tif --F_band 1 -G test.tif --G_band 1 -H test.tif --H_band 1 -I test.tif --I_band 1 -J test.tif --J_band 1 -K test.tif --K_band 1 -L test.tif --L_band 1 -M test.tif --M_band 1 -N test.tif --N_band 1 -O test.tif --O_band 1 -P test.tif --P_band 1 -Q test.tif --Q_band 1 -R test.tif --R_band 1 -S test.tif --S_band 1 -T test.tif --T_band 1 -U test.tif --U_band 1 -V test.tif --V_band 1 -W test.tif --W_band 1 -X test.tif --X_band 1 --outfile=C:\Users\John\AppData\Local\Temp\RtmpK4heMM\file2310254bcc.tif --calc="amin([A,B,C,D,E,F,G,H,I,J,K,L,M,N,O,P,Q,R,S,T,U,V,W,X], axis=0)" --co="COMPRESS=LZW"

0.. evaluation of calculation amin([A,B,C,D,E,F,G,H,I,J,K,L,M,N,O,P,Q,R,S,T,U,V,W,X], axis=0) failed
Traceback (most recent call last):
File "c:\Python33\lib\site-packages\numpy\core\fromnumeric.py", line 2208, in amin
amin = a.min
AttributeError: 'list' object has no attribute 'min'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "C:\Python33\Scripts\GDAL_C~1.PY", line 329, in <module>
main()
File "C:\Python33\Scripts\GDAL_C~1.PY", line 326, in main
doit(opts, args)
File "C:\Python33\Scripts\GDAL_C~1.PY", line 275, in doit
myResult = eval(opts.calc)
File "<string>", line 1, in <module>
File "c:\Python33\lib\site-packages\numpy\core\fromnumeric.py", line 2211, in amin
out=out, keepdims=keepdims)
File "c:\Python33\lib\site-packages\numpy\core\_methods.py", line 29, in _amin
return umr_minimum(a, axis, None, out, keepdims)
ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

同样,在删除对 X 的引用时,它会按照我的预期运行。

我在 Win 10 x64 上使用 Python 3.3.5gdal 1.11.1numpy 1.9.1

为什么会这样?

最佳答案

您遇到的问题是 calc 输入中的 X 变量与 doit 中的循环创建的变量发生冲突> 功能:

for X in range(0,nXBlocks):

这似乎已经被 gdal 开发人员修复(在尚未发布的代码中)。它现在 eval 是私有(private)命名空间(字典)中的 calc 输入,而不是函数的本地命名空间中的 X 变量循环不再与输入的 X 值冲突。

关于python - 传递超过 23 个输入文件时 gdal_calc amin 失败,我们在Stack Overflow上找到一个类似的问题: https://stackoverflow.com/questions/37715316/

25 4 0
Copyright 2021 - 2024 cfsdn All Rights Reserved 蜀ICP备2022000587号
广告合作:1813099741@qq.com 6ren.com