Olá, novamente! Encontrei um erro no código anterior e que está presente também na primeira solução que sugeri. Ao usar {raster} a definição de extent() estava incorreta. Os valores x e y correspondem ao centro do pixel e para o valor de extent() ficar correto deve-se considerar a resolução da imagem. É pouco perceptível na pequena escala, mas pode causar problemas ao aumentar a escala. Segue o código corrigido... ### <code r> ### Dados require(raster) data(volcano); volcano <- 10*volcano image(volcano) x <- 100*(1:nrow(volcano)) y <- 100*(1:ncol(volcano)) ########## {ncdf4} ########################################################### # browseURL("https://www.getdatajoy.com/learn/Read_and_Write_NetCDF_from_R") library(ncdf4) nc1_dim1 <- ncdim_def('EW', 'm', as.double(x)) nc1_dim2 <- ncdim_def('SN', 'm', as.double(y)) nc1_varz <- ncvar_def('Elevation', 'm', list(nc1_dim1, nc1_dim2), -1, longname = 'Data about a Volcano') nc1_out <- nc_create('volcano.nc', nc1_varz, force_v4 = TRUE) ncvar_put(nc1_out, nc1_varz, volcano) nc_close(nc1_out) print(nc_open("volcano.nc")) plot(raster("volcano.nc")) ########## {RNetCDF} ######################################################### # browseURL("https://journal.r-project.org/archive/2013-2/michna-woods.pdf") require(RNetCDF) nc2_out <- create.nc("volcano2.nc") dim.def.nc(nc2_out, "EW", length(x)) dim.def.nc(nc2_out, "SN", length(y)) var.def.nc(nc2_out, "EW", "NC_DOUBLE", "EW") var.def.nc(nc2_out, "SN", "NC_DOUBLE", "SN") var.def.nc(nc2_out, "Elevation", "NC_DOUBLE", c("EW", "SN")) att.put.nc(nc2_out, "Elevation", "_FillValue", "NC_DOUBLE", -999.9) var.put.nc(nc2_out, "EW", as.double(x)) var.put.nc(nc2_out, "SN", as.double(y)) var.put.nc(nc2_out, "Elevation", volcano) close.nc(nc2_out) print(nc_open("volcano2.nc")) ### {ncdf4} print.nc(open.nc("volcano2.nc")) ### {RNetCDF} var.get.nc(nc2, "EW") var.get.nc(nc2, "Elevation") plot(raster("volcano2.nc")) ########## {raster} ########################################################## require(raster) tmp <- raster(volcano) r <- t(flip(tmp, 1)) plot(r) resx <- round(mean(x[-1]-x[-length(x)]), 6) # 100 resy <- round(mean(y[-1]-y[-length(y)]), 6) # 100 extent(r) <- c(range(x), range(y)) + c(-resx, +resx, -resy, +resy)/2 plot(r) writeFormats() writeRaster(r, "volcano3.nc", "CDF", over=T) print(nc_open("volcano3.nc")) ### {ncdf4} plot(raster("volcano3.nc")) r1 <- raster("volcano.nc") r2 <- raster("volcano2.nc") r3 <- raster("volcano3.nc") # extents cbind(sapply(paste0("r", 1:3), function(x) as.vector(extent(get(x))))) # r1 r2 r3 # [1,] 50 50 50 # [2,] 8750 8750 8750 # [3,] 50 50 50 # [4,] 6150 6150 6150 { windows() par(mfrow=c(2,2)) sapply(paste0("r", 1:3), function(x) plot(get(x), main=x)) par(mfrow=c(1,1)) } ### </code> ================================================ Éder Comunello Agronomist (UEM), MSc in Environ. Sciences (UEM) DSc in Agricultural Systems Engineering (USP/Esalq) Brazilian Agricultural Research Corporation (Embrapa) Dourados, MS, Brazil |<O>| ================================================ GEO, -22.2752, -54.8182, 408m UTC-04:00 / DST: UTC-03:00