Provide R-level API and vignette for geographical patterncausality (#868).
Introduce cross mapping of raster data with anisotropic embeddingand variable sliding window steps (#812).
Extend capabilities to incorporate mean-based and stacked spatialcross-sectional embeddings in cross mapping (#800).
Drive extension ofslm generics with customaggregation capabilities (#912).
Permitsimplex andic generics toaccept varying E, k, and tau inputs (#900).
Calibrate spacing conventions and enable real execution ofavailable-thread detection (#885).
Maintain consistent double precision floating-point comparisonsacross cpp sources (#850).
Unify font specification in S3 plotting method for cross-mappingresults (#830).
Support spatial grid(raster) data detrending with cell centercoordinates or row/column numbers (#815).
AdjustMakevars to track latest rcpparmadillorelease and armadillo updates improving build integration and forwardcompatibility (#785).
Remove redundant internal case data (cu.tif) fromthe package (#828).
Harmonize parameter order across s4 generics for spatial vectorand raster data (#817).
Provide new mechanisms for variable interaction in the spatiallogistic map (#772).
Introduce configurable distance metrics for cross mapping(#761).
Enable alternative styles of spatial cross-sectional embedding(#755).
Safeguard transient removal logic in spatial logistic map toprevent index errors (#744).
Adjust the defaultE range in thesimplex generic to2:10 to support more robustreconstruction of state spaces (#739).
Introduce multi-threading in distance-related computations whereapplicable to improve runtime efficiency (#718).
Display p-value annotation for maximum library size in crossmapping visualization legend (#710).
Prevent unreliable predictions and potentially alter results whenNaNs are present in cross mapping (#747).
Rename coordinate columns in population density case csv tolon andlat (#741).
Replace logical vectors with integer index vectors forlib andpred insimplex ands-mapping forecasting sources (#632).
Prevent duplicate generic registrations and simplify methoddefinitions (#627).
Unify parameter descriptions to lowercase (#570).
Introduce confidence interval ribbon support in plot method forgccm results (#550).
Refine randomization strategy in spatial causality test(#643).
SymbolizationC++ functions now compute medians fromlib subset only (#599).
Users must now usedetrend instead oftrend.rm (#559).
Enablecolumn parameter insimplex()andsmap() and renamecolumns parameter tocolumn inmultiview() (#565).
gccm results where legendlabels did not correctly match the corresponding line colors(#552).Enactfnn R API support for false nearest neighboursmethod (#512).
Integrate R API and vignette for geographical cross mappingcardinality (#455).
Provide R-level API and vignette for spatial causality test(#403).
Enable custom legend texts and colors in plot method forgccm results (#535).
Reduce computational load in vignettes (#476).
Document overall structure and usage of spEDM in a dedicatedvignette (#415).
CreateSSR vignette for spatial cross-sectional datastate-space reconstruction (#412).
Include references for algorithms inspEDM(#367).
Use non-NA spatial units forlib/predby default (#499).
Refine internal case data (#348).
Patch memory error caused by mismatch between C++ (0-based) and R(1-based) indexing (#480).
Fix error from non-matrix input in grid-type handling due to Rmatrix slicing (#474).
Enableparallel.level parameter to specify parallelgranularity ingccm R API (#310).
Implement themultiview function for multiviewembedding forecasting (MVE) method (#221).
Integratelib parameter ingccm R APIfor library units selection (#278).
Set the defaultk toE+2 in thegccm R API (#261).
Eliminate redundant computations at the source C++ code level(#233).
Addtrend.rm option in the R API forembedded,simplex, andsmapmethods to align withgccm behavior (#191).
Refactor indexing of lag values and embedding vector generationfor spatial lattice (#186,#184) and grid data(#183,#181).
Default plotting method places the legend in the top-left cornerof the plot now (#325).
Refinesimplex &smap output on theR side (#263).
embedded,simplex,smap when input data contains only one attribute column(#246).sdsfun package(#159).tau parameter in the C++source code and update the R side API (#154).Implement thesmap function to enable the selectionof the optimal theta parameter (#128).
Addsimplex function to support selecting theoptimal embedding dimension for variables (#98).
Provide an R-level API for generating embeddings (#97).
Now bidirectional mapping in thegccm result uses afull join structure when organized on the R side(#118).
Support for calculating unidirectional mappings in thegccm function (#117).
Relaxgccm C++ source codelibsizesminimum value constraint ofE+2 (#109).
Provide a completeGCCM workflow for spatial latticeand grid data in thegccm vignette (#100).
Support testing causal links in GCCM with differentE andk for cause and effect variables(#96).
Add thread settings forgccm generic (#94).
AddS-maps cross-prediction support togccm generic (#81).
Resolve r crash caused by invalidE#90 andk#89 parametersettings ingccm.
Fix incorrect Pearson correlation calculation inC++code when input contains NA (#83).
Encapsulate thegccm generic using the S4 class(#72).
Add options fortau,k, andprogressbar ingccm (#69).
Addprint andplot s3 methods forgccm result (#64).
gccm generic returns empty resultwhen input grid data containsNA values (#61).gccm generic for spatial lattice and griddata using modern C++.