tools::R_user_dir() to a session-specific temporary directory.if(interactive()) with proper \dontrun{} blocks for genuinely non-runable examples (network-dependent, user files required, Suggested packages). CRAN clarified that if(interactive()) does NOT skip examples in checks.\dontrun{} usage to 22 blocks with documented reasons; functions that can run offline use simulated data.set_iobr_cache_dir() or options(IOBR.cache_dir = ...).count2tpm() to return NULL gracefully when annotation data is unavailable offline.r-devel by correcting the http-user-agent configuration and removing redundant dependency specifications.BiocManager to Suggests in DESCRIPTION to improve Bioconductor package resolution and address dependency availability issues reported in CRAN checks.options(IOBR.cache_dir = "your/path"). New functions:
get_iobr_cache_dir() - Get current cache directory pathset_iobr_cache_dir(path) - Set custom cache directoryreset_iobr_cache_dir() - Reset to default system cache location
This enables users to store cached data on shared network drives or any preferred location for easy access across sessions.path / save_path / output.dir is provided.eset_distribution(), find_outlier_samples(), iobr_cor_plot(), sig_pheatmap(), sig_box_batch(), plotPurity(), IPS_calculation(), find_mutations(), sig_gsea(), get_cor(), batch_sig_surv_plot(), format_signatures(), and creat_folder().tempdir() when file writing is demonstrated.find_mutations(): Fixed semantic naming error where file_name variable was used to store directory paths. Renamed to output_dir for clarity. Fixed ggsave() parameter order issues.iobr_cor_plot(): Fixed ggsave() parameter order issues. The correct order is filename first, then plot.surv_group(): Fixed ggsave() parameter order issues.roc_time(): Fixed ggsave() parameter order issues.batch_sig_surv_plot(): Changed default save_path from file.path(tempdir(), "Multiple-KM-plot") to NULL to prevent automatic directory creation.format_signatures(): Changed parameter name from output_name to output_path for consistency. Added validation requiring output_path when save_signature = TRUE.find_outlier_samples(): Added validation requiring project when save = TRUE.plotPurity(): Changed default output.dir from "estimated_purity_plots" to NULL.R/sysdata.rda and data/ to GitHub Releases to meet CRAN package size requirements. Data is now downloaded on-demand and cached locally.load_data() function for unified data access (supports sysdata, exported data, and GitHub-hosted datasets)download_iobr_data() with multiple mirror support (GitHub, ghproxy.vip, gh-proxy.org, ghfast.top)add_iobr_mirror() for custom mirror configurationlist_github_datasets() to show available remote datasetsclear_iobr_cache() to manage downloaded datacount2tpm() to handle symbol-based gene IDs more robustly.The following datasets are now hosted on GitHub and downloaded on first use:
BRef, TRef, lm22anno_gc_vm32, anno_grch38, anno_hug133plus2, anno_illumina, anno_rnaseqtcga_stad_sig, imvigor210_sig, eset_stad, sig_stad, eset_gse62254, etc.hallmark, kegg, go_bp, go_cc, go_mf, reactome, msig_immune, msig_sccancer_type_genes, cellmarkers, common_genes, immuneCuratedData, ips_gene_set, SI_geneset, mRNA_cell_default, mus_human_gene_symbol, onco_sig, PurityDataAffyxCell.data, quantiseq_datasignature_collection_citation, signature_metabolism, signature_sc, signature_tumorload_data() instead of direct object references for migrated datasetsglobalVariables.R to reflect data migrationghcr.io/iobr/iobr). The image is based on rocker/tidyverse and includes IOBR with all dependencies pre-installed (#Docker).iobr_cor_plot() to support multi-group (3+) comparisons using Kruskal-Wallis test, in addition to the existing two-group Wilcoxon test (#28).generateRef_DEseq2() to handle sparse single-cell RNA-seq data by using type = "poscounts" for size factor estimation, preventing errors when all genes contain at least one zero.batch_kruskal() output format to correctly display mean-centered values for all groups.count2tpm() computation results to align with previous correct versions.count2tpm() documentation to clarify that gene identifiers are converted to gene symbols in the output regardless of input ID type.\dontrun{} and \donttest{} to comply with CRAN check time limits.count2tpm.R for better code organization and maintainability.load_data() documentation to include additional available datasets.