Skip to contents

Changelog

Source: NEWS.md

ckm 1.1.0

Add the build_stacked_matrix function to let user apply a stacked matrix as presented in Jamme (2026) doi:10.13140/RG.2.2.19275.55840. This is an effective way to improve the risk-utility trade-off, especially when user has to exclude small counts to lower the risk on them. The stacked matrix lets lower this risk by monitoring the loss on the other counts.

To use this new feature, the functions apply_ckm, tabulate_and_apply_ckm and tabulate_and_apply_ckm_list propose the stack argument. By default NULL, the user can activate the stacked matrix by setting this argument with the list of the parameters to apply only on the “big” counts. For example, if the user set the following parameters : D=15, V=35, js=10, stack = list(D=15, V=5), then the transition matrix used to apply the Cell Key Method is built with the first three for all the rows except the last one, and the parameters D=15, V=5, js=0 will be used on the last row.

ckm 1.0.2

Improve estimation of frequencies in compute_frequencies.

ckm 1.0.1

Original names of the functions have been translated to English:

  • construire_cles_indiv -> build_individual_keys
  • tabuler_et_appliquer_ckm -> tabulate_and_apply_ckm
  • appliquer_ckm -> apply_ckm
  • creer_matrice_transition -> create_transition_matrix
  • creer_table_perturbation -> create_perturbation_table
  • tester_matrices -> test_matrices
  • mesurer_risque -> assess_risk
  • mesurer_RU -> assess_RU
  • mesurer_RUs -> assess_RUs
  • simuler_RUs -> simulate_RUs
  • tabuler_et_appliquer_ckm_liste -> tabulate_and_apply_ckm_list
  • visualiser_distribution -> visualize_distribution
  • estimer_beta -> estimate_beta
  • estimer_beta_df -> estimate_beta_df

ckm 0.3.9410

Modification:

  • Possibility to measure the risk and utility of one or more scenarios:
    • measure_RU: Measures the risk and utility of a given scenario
    • measure_RUs: Compares the risk and utility of several scenarios
    • simulate_RUs: Compares the risk and utility of several scenarios based on multiple simulations
  • By default, the arguments I and J are set to NULL, disabling risk measurement by default when using apply_ckm or tabuler_et_appliquer_ckm.

New function:

  • Apply the method to several tables at once: tabulate_and_apply_ckm_list

ckm 0.2.2

[13/11/2024]

Modification:

  • tester_matrices: la fonction recherche désormais une variance minimale dans un intervalle défini par l’utilisateur et par dichotomie

ckm 0.2.0

[29/10/24]

Ajout:

  • mesure du risque
  • mesure de l’utilité
  • recherche variance

Modification:

  • construire_cles_indiv: La dépendance au package cellKey est supprimée et l’argument seed a été supprimé. L’utilisateur devra appeler la fonction set.seed() pour fixer la graine aléatoire.
  • appliquer_ckm et tabuler_et_appliquer_ckm revues pour intégrer la mesure du risque et de l’utilité.

ckm 0.1.0

[02/08/24]

Première version contenant les fonctions :

  • construire_cles_indiv
  • tabulate_cnt_micro_data
  • tabuler_et_appliquer_ckm
  • appliquer_ckm
  • creer_matrice_transition
  • creer_table_perturbation