sits_labelssits_trainroi parameter insits_mosaic andsits_plotsits_accuracy messages when results areemptyTAE implementation to make better use ofembeddingssits_cube_copysits_textureres parameter insits_mosaicsits_roi_to_tiles functionsits_get_data() implementationsits_mosaic()sits_clean() multicores operationssits_view() usingleafglsits_summary() andsits_stratified samplingsits_regularize()sits_select()exclusion_mask parameter insits_classify() andsits_smooth()sits_regularize(), including MGRS and Brazil Data Cubegridssits_merge() implementation to better handlemultiple scenario casesroi when plotting data cubessits_cube_copy() to improve timeout handlingand efficiencysits_list_collections()SpatExtent object from terra asroi insits_cube()crs usage insits_get_data() tosupport WKTsits_classify() performance with segmentsclassification.reg_cube_split_assets() for R 4.Xcompatibilitysits_merge() function that was not mergingSAR andOPTICAL cubessits_view()plot() performance using raster overviewssits_cube()sits_mosaic()sits_segment() using chunkparallelizationsits_clean() function to improve classifiedmapssits_sampling_design() andsits_stratified_sampling()sits_reduce() functiondtw distance when building SOM mapssits_classify()segmentssits_apply()supercellspackagesits_get_data() to extract averagevalues of time series based on segmentssits_view()summary() function to show details of data cubesand time series tibblessits_mosaic() function for improving visualizationof large data setssits_regularize()sits_cube_copy() for downloading data from theinternetsitssits_train()sits_combine_predictions()data.table package.raster_file_blocksize.terra() bug (issue#918)stars proxy bug (issue #902)purrr cross deprecationggplot2 aes_string deprecationtibble subsetting bug (issue #893)sits_som_clean_samples() bug (issue #890)sits_get_data() can be used to retrieve samples inclassified cubesits_mixture_model())sits_mosaic_cubes())sits_model())sits_cube_copy())sits_combine_predictions())sits_plot)sits_apply()sits_regularize() (issue#848)sits_labels()<- (issue #846)sits_label_classification()andsits_smooth() (issue #850)sits_classify() on BDC cubes(issue #844)sits_apply()sits_apply()sits_applysits_mixture_model for spectralmixture analysissits_viewsits_as_sf to convert sitsobjects to sfsits_regularizeroi parameter insits_regularizefunctioncrs parameter insits_get_data"MPC"sits_whittaker() function to processcube.sits_lighttae()(Lightweight Temporal Self-Attention)sits_uncertainty_sampling() for activelearningsits_confidence_samples() forsemi-supervised learningsits_geo_dist() to generate samples-samplesand samples-predicted plotsits_tuning() for random search of machinelearning parameterssits_reduce_imbalance() function to balanceclass samplessits_as_sf() to convert a sits tibble to asf objecttorchopt deep learning optimizerpackagesits_uncertainty():leastconfidence andmargin of confidencesits_kfold_validate()data tosamples in sits machinelearning classifiers (NOTE: models trained in previous versions is nolonger supported)file parameter insits_get_data()functiontorchpackage and removekeras dependencesits_TAE() classification modelsits_lightgbm() classification modelsits_regularize() parameterssits_regularize() to reach production levelqualitysits_regularize() to use C++ internalfunctionssits_cube() to open results cubeplot() parameters on raster cubessits_view()sits_get_data() to accept tibblessits_cube()sits_regularize() to process in parallel bytiles, bands, and datessits_regularize() to check malformed filesAWS_NO_SIGN_REQUEST environment variable.gc_get_valid_interval() function.sits_regularize has a fault tolerance system, sothat if there is a processing error the function will delete themalformed files and create them again.sits_regularize function has a new parameter calledmultithreads.sits_cube function forlocal cubes has anew parameter calledmulticores.F1 score insits_kfold_validate withmore than 2 labels.sits_cube() function to tolerate malformed pathsfrom STAC service;sits_apply() function to generate new bandsfrom existing ones;sits_accuracy() function to work with multiplecubes;sits_view()sits_uncertainty() function to provideuncertainty measure to probability maps;sits_regularize() by taking least cloud coverby default method to compose imagessits_regularize that generated images withartifactssits_cube from STAC AWSSentinel-2sits_timeline() to sits model objectsconfig_colors.yml by removing palettenamessits_regularize()start_date andend_date fromvalidation csv filesits_regularize() is producingFloat64 imagesas outputgdalcubes_chunk_size in “config.yml” to improvesits_regularize()..source_collection_access_test to passellipsis torstac::post_request function..source_collection_access_test to passellipsis torstac::post_request function.sits_plotsits_timeline for cubes that do not have thesame temporal extent.S2_10_16D_STK-1 removed from BDC source inconfig fileNoClass label improvementmapview toleaflet packageCLASSIFIED andPROBS sources fromconfig fileterra package to1.4-11sits_list_collections() to indicate open datacollectionptw,signal andMASSopen_data collections in configfileoutput_dir parametersits_cube_clone() functionsits_select() for bands in rastercubesits_regularize()functionOPENDATA sourceS2_10-1 BDC collection from configsits_list_collections().source_bands_resampling()sits_som_clean_samples() functionsits_bands<-() functionsits_select() functionsits_bbox() functionS2-SEN2COR_10_16D_STK-1 BDC collectioncheck functionsatellite andsensor info in configfileimager,ranger,proto,andfuture packages from sitssits_cube.local_cube() function parameterssatellite andsensororigin andcollection tosits_cube.local_cube() functionroi parameter insits_classify() functionRaster classification results can now have versions: a newparameter “version” has been included in thesits_classifyfunction.
Corrections tosits_kohonen and to thedocumentation.
New deep learning models for time series: 1D convolutional neuralnetworks (sits_FCN), combining 1D CNN and multi-layerperceptron networks (sits_TempCNN), 1D version of ResNet(sits_ResNet), and combination of long-short term memory(LSTM) and 1D CNN (sits_LSTM_FCN).
New version of area accuracy measures that include Olofssonmetrics ()
From version 0.8 onwards, the package has been designed to workwith data cubes. All references to “coverage” have been replaced byreferences to “cubes”.
The classification of raster images usingsits_classify now produces images with the information onthe probability of each class for each pixel. This allows moreflexibility in the options for labeling the resulting probability rasterfiles.
The functionsits_label_classification has beenintroduced to generate a labelled image from the class probabilityfiles, with optional smoothing. The choices aresmoothing = none (default),smoothing = bayesian (for bayesian smoothing) andsmoothing = majority (for majority smoothing).
To better define a cube, the metadata tibble associated to a cuberequires four parameters to define the cube: (a) the web service thatprovides time series or cubes; (b) the URL of the web service; (c) thename of the satellite; (d) the name of the satellite sensor. If notprovided, these parameters are inferred for thesitsconfiguration file.
The functions that do data transformations, such assits_tasseled_cap andsits_savi now require asensor parameter (“MODIS” is the default)
Functionssits_bands andsits_labelsnow work for both tibbles with time series and data cubes.
sits_show_config() to see the default contents. Userscan override these parameters or add their own by creating aconfig.yml file in their home directory.Examples and demos that include classification of raster filesnow use theinSitu R package, available usingdevtools::install_github(e-sensing/inSitu).
All examples have been tested and checked forcorrectness.
sits_coverage has been replaced bysits_cube.
sits_raster_classification has been removed. Pleaseusesits_classify.
Insits_classify, the parameterout_prefix has been changed tooutput_dir, toallow better control of the directory on which to write.
sits_bayes_smooth has been removed. Please usesits_label_classification withsmoothing = bayesian.
To define a cube based on local files,service = RASTER has been replaced byservice = LOCALHOST.
For programmers only: Thesits_cube.R file nowincludes many convenience functions to avoid using cumbersome indexes tofiles and vector:.sits_raster_params,.sits_cube_all_robjs,.sits_class_band_name,.sits_cube_bands,.sits_cube_service,.sits_cube_file,.sits_cube_files,.sits_cube_labels,.sits_cube_timeline,.sits_cube_robj,.sits_cube_all_robjs,.sits_cube_missing_values,.sits_cube_minimum_values,.sits_cube_maximum_values,.sits_cube_scale_factors,.sits_files_robj.Please look at the documentation provided in thesits_cube.R file.
For programmers only: The metadata that describes the data cubeno longer stores the raster objects associated to the files associatedwith the cube.