Compind v3.8 (Release date: 2026-11-19) ============== Changes: * ci_rbod_constr_bad_Q: removed the unused arguments cykertype, cxkertype and oxkertype from both internal npksum() calls, for compatibility with np >= 0.80-3. Corrected bugs: * ci_mean_min: standardisation is now column-wise (the column means and standard deviations were recycled by row). * ci_smaa_constr: each draw of weights now sums to one and respects the lower bounds (low_w); weights are drawn uniformly on the constrained simplex. * bandwidth_CI, ci_rbod_constr_bad_Q: ordered exogenous variables (Q_ord) are now passed to np as ordered factors, with levels fixed on the full sample. * bandwidth_CI: corrected the selection of the ordered columns when both Q and Q_ord are supplied. * ci_geom_gen, ci_geom_bod_intertemp: indic_col is no longer applied twice (results were wrong or the function failed when indic_col did not start at 1). * ci_rbod_constr_bad, ci_rbod_constr_bad_Q: no longer fail with a preference vector (pref) of two elements and more than two indicators. * ci_bod_constr_bad, ci_rbod_constr_bad, ci_rbod_constr_bad_Q: a preference vector with a single element now stops with an error (results were silently meaningless). Faster code (same results): * ci_ogwa, * ci_owa * ci_rbod_dir * ci_bod_constr * ci_bod_constr_bad * ci_rbod_constr_bad * ci_rbod_constr_bad_Q * ci_rbod_mdir * bandwidth_CI Compind v3.7 (Release date: 2026-10-06) ============== Changes: Corrected bugs for * ci_generalized_mean * ci_rbod_dir * ci_rbod_mdir * plot_M_robust_ci Compind v3.6 (Release date: 2026-07-10) ============== Changes: New functions: * ci_bod_time_decomp.R : Dynamic Benefit of the Doubt decomposition for panel data Compind v3.4 (Release date: 2025-05-15) ============== Changes: New functions: * plot_M_robust_ci : Visualize the percentage of outperforming units in robust methods * ci_rbod_mdir: Internally improved function Compind v3.3 (Release date: 2025-05-9) ============== Changes: New functions: * ci_mpi.R: ATTENTION: This function has been rewritten and it is therefore necessary to normalise simple data BEFORE using the function; see the example in the help. Compind v3.2 (Release date: 2025-1-9) ============== Changes: * ci_ampi: corrected internal errors Compind v3.1 (Release date: 2024-03-25) ============== Changes: New functions: * ci_rbod_mdir: Robust Multi-directional Benefit of the Doubt approach (MDRBoD) Compind v3.0 (Release date: 2023-12-28) ============== Changes: New functions: * ci_owa: Ordered Weighted Average (OWA) * ci_ogwa: Ordered Geographically Weighted Average (OGWA) Compind v2.9 (Release date: 2023-11-14) ============== Changes: New function: * ci_factor_mixed: Weighting method based on Factor analysis of mixed data (FAMD) Compind v2.8 (Release date: 2023-08-20) ============== Changes: New function: * ci_bod_mdir: Multi-directional Benefit of the Doubt approach (MDBoD) Compind v2.7 (Release date: 2023-01-19) ============== Changes: * Some bugs have been fixed. Compind v2.6 (Release date: 2022-7-05) ============== Changes: * Solved problem in a linked package (nonparaeff). function ci_bod_var_w is now at your disposal Compind v2.5 (Release date: 2022-6-20) ============== Changes: * problem in a linked package (nonparaeff). Removed function ci_bod_var_w Compind v2.4 (Release date: 2022-5-24) ============== Changes: * problem in a linked package (smaa). This problem has been fixed. Compind v2.3 (Release date: 2021-4-09) ============== Changes: * ci_smaa_constr function example corrected Compind v2.2 (Release date: 2021-4-9) ============== Changes: * ci_ampi (Adjusted Mazziotta-Pareto Index (AMPI) method) function fixed: (1) negative penalty as default; (2) fixed, in the help, the explanation of the penalty that is = NEG for positive indicators and = POS for negative indicators (for example: deprivation); (3) calculation of standard deviation divided by n (and not by n-1). Compind v2.0 (Release date: 2018-2-13) ============== Changes: New functions: * ci_bod_constr_bad: Constrained Benefit of the Doubt approach (BoD) in presence of undesirable indicators * ci_rbod_constr_bad: Robust constrained Benefit of the Doubt approach (BoD) in presence of undesirable indicators. * ci_rbod_constr_bad_Q: Conditional robust constrained Benefit of the Doubt approach (BoD) in presence of undesirable indicators * ci_rbod_spatial: Spatial robust Benefit of the Doubt approach (Sp-RBoD) * ci_smaa_constr: Constrained stochastic multi-objective acceptability analysis (C-SMAA) * ci_ampi: Adjusted Mazziotta-Pareto Index (AMPI) method Compind v1.2.1 (Release date: 2017-07-12) ============== Changes: * ci_geom_bod_intertemp: correction due to new R functionalities Compind v1.2 (Release date: 2017-16-06) ============== Changes: New functions: * ci_bod_constr: Constrained Benefit of the Doubt approach * ci_generalized_mean: Weighting method based on generalized mean * ci_geom_bod_intertemp: Intertemporal analysis for geometric mean quantity index numbers * ci_geom_gen: generalization of the previous function "ci_mean_geom" Deleted * ci_bod_vrs Compind v1.1.2 (Release date: 2016-06-27) ============== Changes: * Compind::Compind_vignette has been included. Compind v1.1 (Release date: 2015-09-07) ============== Changes: * ci_factor function: new method (CH), modified old function * ci_bod : Bod weights are avalaible * ci_mean_min: Mean-Min Function (MMF) .