Pre-release 1.4.2-1

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: sits
 Type: Package
-Version: 1.4.2
+Version: 1.4.2-1
 Title: Satellite Image Time Series Analysis for Earth Observation Data Cubes
 Authors@R: c(person('Rolf', 'Simoes', role = c('aut'), email = 'rolf.simoes@inpe.br'),
              person('Gilberto', 'Camara', role = c('aut', 'cre'), email = 'gilberto.camara.inpe@gmail.com'),
@@ -36,7 +36,7 @@ Description: An end-to-end toolkit for land use and land cover classification
     Minimum recommended requirements: 16 GB RAM and 4 CPU dual-core.
 Encoding: UTF-8
 Language: en-US
-Depends: R (>= 4.1.0)
+Depends: R (>= 4.0.0)
 URL: https://github.com/e-sensing/sits/, https://e-sensing.github.io/sitsbook/
 BugReports: https://github.com/e-sensing/sits/issues
 License: GPL-2

NEWS.md

@@ -2,6 +2,12 @@
 
 # What's new in SITS version 1.4
 
+### Hotfix version 1.4.2-1
+* Fix crs bug in `sits_apply()`
+* Update file name in clean feature
+* Fix time series extraction bug with segments
+* Fix examples
+
 ### New features in SITS version 1.4.2
 * Support for vector data cubes, including visualisation
 * Object-based time series analysis using spatio-temporal segmentation

R/api_clean.R

@@ -23,25 +23,9 @@
                         output_dir,
                         version) {
     # Output file
-    out_file <- .file_clean_name(
-        tile = tile, band = band,
-        version = version, output_dir = output_dir
+    out_file <- .file_derived_name(
+        tile = tile, band = band, version = version, output_dir = output_dir
     )
-    # Resume asset
-    if (.raster_is_valid(out_file, output_dir = output_dir)) {
-        # recovery message
-        .check_recovery(out_file)
-        # Create tile based on template
-        tile <- .tile_derived_from_file(
-            file = out_file, band = band,
-            base_tile = tile, derived_class = .tile_derived_class(tile),
-            labels = .tile_labels(tile),
-            update_bbox = FALSE
-        )
-        return(tile)
-    }
-    # Remove remaining incomplete files
-    unlink(out_file)
     # Create chunks as jobs
     chunks <- .tile_chunks_create(
         tile = tile, overlap = overlap, block = block
@@ -88,6 +72,8 @@
         # Returned block files for each fraction
         block_files
     })
+    # Remove raster without clean
+    unlink(.tile_path(tile))
     # Merge blocks into a new class_cube tile
     band_tile <- .tile_derived_merge_blocks(
         file = out_file,

R/api_file.R

@@ -119,22 +119,6 @@
     )
 }
 
-#' @title Build a file path for a cleaned map (used in sits_clean)
-#' @noRd
-#' @param tile     Tile of data cube
-#' @param band     Spectral band
-#' @param version Version name
-#' @param output_dir Directory where file will be saved
-#' @returns        File path for derived cleaned map
-.file_clean_name <- function(tile, band, version, output_dir) {
-    .file_path(
-        tile[["satellite"]], tile[["sensor"]], .tile_name(tile),
-        .tile_start_date(tile), .tile_end_date(tile), band,
-        paste(version, "clean", sep = "-"),
-        ext = "tif", output_dir = output_dir
-    )
-}
-
 #' @title Build a file path for a mosaic
 #' @noRd
 #' @param tile     Tile of data cube

R/api_file_info.R

@@ -57,10 +57,11 @@ NULL
 #' @param xmax largest X coordinate
 #' @param ymin smallest Y coordinate
 #' @param ymax largest Y coordinate
+#' @param crs  coordinate reference system
 #' @param path location of the data
 #' @returns  eo_cube tibble
 .fi_eo <- function(fid, band, date, ncols, nrows, xres, yres, xmin, xmax,
-                   ymin, ymax, path) {
+                   ymin, ymax, crs, path) {
     # Create a new eo file_info
     tibble::tibble(
         fid = fid,
@@ -74,6 +75,7 @@ NULL
         xmax = xmax,
         ymin = ymin,
         ymax = ymax,
+        crs = crs,
         path = path
     )
 }
@@ -99,6 +101,7 @@ NULL
         xmax = .raster_xmax(r_obj),
         ymin = .raster_ymin(r_obj),
         ymax = .raster_ymax(r_obj),
+        crs = .raster_crs(r_obj),
         path = files
     )
 }

R/api_label_class.R

@@ -12,19 +12,6 @@
     out_file <- .file_derived_name(
         tile = tile, band = band, version = version, output_dir = output_dir
     )
-    # Resume feature
-    if (file.exists(out_file)) {
-        .check_recovery(tile[["tile"]])
-        class_tile <- .tile_derived_from_file(
-            file = out_file,
-            band = band,
-            base_tile = tile,
-            derived_class = "class_cube",
-            labels = .tile_labels(tile),
-            update_bbox = FALSE
-        )
-        return(class_tile)
-    }
     # Create chunks as jobs
     chunks <- .tile_chunks_create(tile = tile, overlap = 0)
     # Process jobs in parallel

R/api_segments.R

@@ -271,6 +271,13 @@
                             time_series = c("Index", dplyr::all_of(bands)))
     # include the ids of the polygons
     ts_bands[["polygon_id"]] <- pol_id
+    # we do the unnest again because we do not know the polygon id index
+    ts_bands <- tidyr::unnest(ts_bands, "time_series")
+    # remove pixels where all timeline was NA
+    ts_bands <-  tidyr::drop_na(ts_bands)
+    # nest the values by bands
+    ts_bands <- tidyr::nest(ts_bands,
+                            time_series = c("Index", dplyr::all_of(bands)))
     # nest the values by sample_id and time_series
     ts_bands <- tidyr::nest(ts_bands, points = c("sample_id", "time_series"))
     # retrieve the segments
@@ -418,12 +425,12 @@
             output_dir = output_dir)
         samples <- purrr::map(seg_tile_bands_lst, function(seg_tile_bands){
             samples_part <- .segments_extract_data(seg_tile_bands,
-                                              tile,
-                                              bands,
-                                              pol_id,
-                                              n_sam_pol,
-                                              multicores,
-                                              progress)
+                                                   tile,
+                                                   bands,
+                                                   pol_id,
+                                                   n_sam_pol,
+                                                   multicores,
+                                                   progress)
             return(samples_part)
         })
         samples <- dplyr::bind_rows(samples)

R/api_tile.R

@@ -1182,7 +1182,7 @@ NULL
 #' @param derived_class class of the derived tile
 #' @param block_files  files that host the blocks
 #' @param multicores  number of parallel processes
-#' @param update_bbox   should bbox be updated?
+#' @param update_bbox should bbox be updated?
 #' @return a new tile with files written
 .tile_derived_merge_blocks <- function(file, band, labels, base_tile,
                                        derived_class, block_files, multicores,

R/sits_apply.R

@@ -133,6 +133,14 @@ sits_apply.raster_cube <- function(data, ...,
     # Get output band expression
     expr <- .apply_capture_expression(...)
     out_band <- names(expr)
+    # Check if band already exists in cube
+    if (out_band %in% .cube_bands(data)) {
+        warning(
+            paste0("The provided band '", out_band, "' already exists in cube."),
+            call. = FALSE
+        )
+        return(data)
+    }
     # Get all input bands in cube data
     in_bands <- .apply_input_bands(data, expr = expr)
 

R/sits_classify.R

@@ -133,14 +133,16 @@
 #'         ml_model = rf_model,
 #'         output_dir = tempdir(),
 #'         n_sam_pol = 20,
-#'         multicores = 4
+#'         multicores = 4,
+#'         version = "segs_classify"
 #'     )
 #'     # Create a labelled vector cube
 #'     class_segs <- sits_label_classification(
 #'         cube = probs_segs,
 #'         output_dir = tempdir(),
 #'         multicores = 2,
-#'         memsize = 4
+#'         memsize = 4,
+#'         version = "segs_classify"
 #'     )
 #'     # plot class_segs
 #'     plot(class_segs)

R/sits_label_classification.R

@@ -120,6 +120,24 @@ sits_label_classification.probs_cube <- function(cube, ...,
     label_fn <- .label_fn_majority()
     # Process each tile sequentially
     class_cube <- .cube_foreach_tile(cube, function(tile) {
+        # Output file
+        out_file <- .file_derived_name(
+            tile = tile, band = "class", version = version,
+            output_dir = output_dir
+        )
+        # Resume feature
+        if (file.exists(out_file)) {
+            .check_recovery(tile[["tile"]])
+            class_tile <- .tile_derived_from_file(
+                file = out_file,
+                band = "class",
+                base_tile = tile,
+                derived_class = "class_cube",
+                labels = .tile_labels(tile),
+                update_bbox = FALSE
+            )
+            return(class_tile)
+        }
         # Label the data
         class_tile <- .label_tile(
             tile = tile,
@@ -129,7 +147,6 @@ sits_label_classification.probs_cube <- function(cube, ...,
             version = version,
             progress = progress
         )
-        label_path <- .tile_path(class_tile)
         if (clean) {
             # Apply clean in data
             class_tile <- .clean_tile(
@@ -141,8 +158,6 @@ sits_label_classification.probs_cube <- function(cube, ...,
                 output_dir = output_dir,
                 version = version
             )
-            # Remove raster file without clean
-            unlink(label_path)
         }
         return(class_tile)
     })

R/sits_plot.R

@@ -1380,8 +1380,6 @@ plot.som_map <- function(x, y, ..., type = "codes", band = 1) {
 #'     xgb_model <- sits_train(samples_modis_ndvi,
 #'         ml_method = sits_xgboost()
 #'     )
-#'     # plot the model
-#'     plot(xgb_model)
 #' }
 #' @export
 #'

R/sits_sample_functions.R

@@ -229,6 +229,8 @@ sits_reduce_imbalance <- function(samples,
         )
         new_samples <- dplyr::bind_rows(new_samples, samples_classes_ok)
     }
+    # remove SOM additional columns
+    colnames_sits <- setdiff(colnames(new_samples), c("id_neuron", "id_sample"))
     # return new sample set
-    return(new_samples)
+    return(new_samples[, colnames_sits])
 }

R/sits_segmentation.R

@@ -209,20 +209,23 @@ sits_segment <- function(cube,
 #'     # segment the vector cube
 #'     segments <- sits_segment(
 #'         cube = cube,
-#'         output_dir = tempdir()
+#'         output_dir = tempdir(),
+#'         version = "slic-demo"
 #'     )
 #'     # create a classification model
 #'     rfor_model <- sits_train(samples_modis_ndvi, sits_rfor())
 #'     # classify the segments
 #'     seg_probs <- sits_classify(
 #'         data = segments,
 #'         ml_model = rfor_model,
-#'         output_dir = tempdir()
+#'         output_dir = tempdir(),
+#'         version = "slic-demo"
 #'     )
 #'     # label the probability segments
 #'     seg_label <- sits_label_classification(
 #'         cube = seg_probs,
-#'         output_dir = tempdir()
+#'         output_dir = tempdir(),
+#'         version = "slic-demo"
 #'     )
 #' }
 #' @export

README.Rmd

@@ -94,12 +94,16 @@ devtools::install_github("e-sensing/sits", dependencies = TRUE)
 library(sits)
 ```
 
+### Support for GPU 
+
+Classification using torch-based deep learning models in `sits` uses CUDA compatible NVIDIA GPUs if available, which provides up 10-fold speed-up compared to using CPUs only. Please see the [installation instructions](https://torch.mlverse.org/docs/articles/installation) for more information on how to install the required drivers. 
+
 
 ## Building Earth Observation Data Cubes
 
 ### Image Collections Accessible by `sits`
 
-The `sits` package allows users to created data cubes from analysis-ready data (ARD) image collections available in cloud services. The collections accessible in `sits` `r packageVersion("sits")` are:
+Users create data cubes from analysis-ready data (ARD) image collections available in cloud services. The collections accessible in `sits` `r packageVersion("sits")` are:
 
 1. Brazil Data Cube ([BDC](http://brazildatacube.org/en/home-page-2/#dataproducts)): Open data collections of Sentinel-2, Landsat-8 and CBERS-4 images.
 2. Microsoft Planetary Computer ([MPC](https://planetarycomputer.microsoft.com/catalog)): Open data collection of Sentinel-2/2A and Landsat-8
@@ -273,7 +277,9 @@ plot(label_cube,
 
 ## Additional information
 
-For more information, please see the on-line book ["SITS: Data analysis and machine learning for data cubes using satellite image time series"](https://e-sensing.github.io/sitsbook/).
+Since version 1.4.2, `sits` support OBIA analysis of image time series, using an extension of R package `supercells`.  
+
+The package is described in detail in on-line book ["SITS: Data analysis and machine learning for data cubes using satellite image time series"](https://e-sensing.github.io/sitsbook/).
 
 
 ### References
@@ -314,11 +320,15 @@ We thank the authors of these papers for making their code available to be used
 
 - [12] Maja Schneider, Marco Körner, "[Re] Satellite Image Time Series Classification with Pixel-Set Encoders and Temporal Self-Attention." ReScience C 7 (2), 2021. <doi:10.5281/zenodo.4835356>.
 
-#### R packages used in sits
+- [13] Jakub Nowosad, Tomasz Stepinski, "Extended SLIC superpixels algorithm for applications to non-imagery geospatial rasters". International Journal of Applied Earth Observation and Geoinformation, 112, 102935, 2022.
+
+- [14] Martin Tennekes, “tmap: Thematic Maps in R.” Journal of Statistical Software, 84(6), 1–39, 2018.
+
+### Acknowledgements for community support
 
-The authors are thankful for the contributions of Marius Appel, Tim Appelhans, Henrik Bengtsson, Robert Hijmans, Edzer Pebesma, and Ron Wehrens, respectively chief developers of the packages `gdalcubes`, `leafem`, `data.table`, `terra/raster`, `sf`/`stars`, and `kohonen`. The `sits` package  is also much indebted to the work of the RStudio team, including the `tidyverse`. We are indepted to Daniel Falbel for his and the `torch` packages. We thank Charlotte Pelletier and Hassan Fawaz for sharing the python code that has been reused for the TempCNN and ResNet machine learning models. We would like to thank Maja Schneider for sharing the python code that helped the implementation of the `sits_lighttae()` and `sits_tae()` model. We recognise the importance of the work by Chris Holmes and Mattias Mohr on the STAC specification and API. 
+The authors are thankful for the contributions of Edzer Pebesma, Jakub Novosad. Marius Appel, Martin Tennekes, Robert Hijmans, Ron Wehrens, and Tim Appelhans, respectively chief developers of the packages `sf`/`stars`, `supercells`, `gdalcubes`, `tmap`, `terra`, `kohonen`, and  `leafem`. The `sits` package  is also much indebted to the work of the RStudio team, including the `tidyverse`. We are indepted to Daniel Falbel for his great work in the `torch` and `luz` packages. We thank Charlotte Pelletier and Hassan Fawaz for sharing the python code that has been reused for the TempCNN and ResNet machine learning models. We would like to thank Maja Schneider for sharing the python code that helped the implementation of the `sits_lighttae()` and `sits_tae()` model. We recognise the importance of the work by Chris Holmes and Mattias Mohr on the STAC specification and API. 
 
-## Acknowledgements for Financial and Material Support
+### Acknowledgements for Financial and Material Support
 
 We acknowledge and thank the project funders that provided financial and material support:
 
@@ -336,7 +346,7 @@ We acknowledge and thank the project funders that provided financial and materia
     funding from the European Union's Horizon Europe research and innovation programme 
     under [grant agreement No. 101059548](https://cordis.europa.eu/project/id/101059548). 
 
-## How to contribute
+### How to contribute
 
 The `sits` project is released with a [Contributor Code of Conduct](https://github.com/e-sensing/sits/blob/master/CODE_OF_CONDUCT.md).
 By contributing to this project, you agree to abide by its terms.