Distributacalcul 0.3.1 Unreleased

Bug fixes

  • Changed location of translations file to not be ignored when built and and published.
  • Should avoid issue I’m getting running the visualisation functions.

Distributacalcul 0.3.0 2020-09-11

Significant structural change

  • Changed syntax of all package functions (see README for summary of new syntax).
    • As per comments of peers, the syntax was inconsistent.
    • Changed all functions names to follow lowerCamelCase.
    • E_d -> expValD; V_d -> varD; Elim_d -> expValLimD; Etrunc_d -> expValTruncD; Mexcess_d -> meanExcessD; SL_d -> stopLossD; VaR_d -> VatRD; TVaR_d -> TVatRD; kthmoment_d -> kthMomentD; MGF_d -> mgfD; PGF_d -> pgfD; p_d -> pD; d_d -> dD.
  • Added MGF for the Uniform distribution.
  • Combined all separate function files to files containing the entire distribution.
    • Amongst other things, it simplifies working with the package.
  • Added template files for the parameters which didn’t already have some.
    • dispersion parameter; mean parameters for the Normal (mu), lognormal, and Inverse Gaussian (location) distributions; min/max parameters for the Uniform distributions; number of tries parameter for the Negative Binomial; prob parameter of the logarithmic distribution; rate as the normal beta, as an alternative parameter and as lambda; scale parameter as normal and as lambda; standard deviation parameter for both the normal and lognormal distributions; shape parameter as the normal alpha, the n for the Erlang, and as tau; shape1 and shape2 for the Beta.
  • Changed name of compound distributions as well as the others.
  • Changed negative binomial functions names from negbinom to nbinom.

Significant New Features

  • Added copulas !
    • copula functions cCopula, density functions associated with the copulas cdCopula, and simulation functions crCopula.
  • Added copulas are :
    1. Independence Copula
    2. Fréchet Lower Bound Copula
    3. Fréchet Upper Bound Copula
    4. Fréchet Copula
    5. Bivariate Gumbel Copula
    6. Bivariate Clayton Copula
    7. Bivariate Ali-Mikhail-Haq Copula
    8. Bivariate Cuadras-Augé Copula
    9. Bivariate Marshall-Olkin Copula
    10. Bivariate Frank Copula
    11. Bivariate Eyraud-Farlie-Gumbel-Morgenstern (EFGM) Copula
  • Added visualisation function copulaVis for Copulas.
    • Internal functions are distributionPlotBoxCopulas (3D plot of distribution), simulationPlotBoxCopulas (hexagon plot of simulated observations) and parametersBoxCopulas (parameters box).

New Features

  • Added pllogis() and dllogis() functions.
    • Added the functions to the return field of the loglogistic-template file.
  • Added plogarithmic(), dlogarithmic(), and VaR_logarithmic()functions.
    • Added the functions to the return field of the logarithmic-template file.
  • Created vignette distributionParameters:
    • Explains the various parameters of distributions (scale, rate, location).
    • Still need to add shape and dispersion.
    • Still need to fix viewing of vignette.
  • Added functionality to MGF_negbinom() and PGF_negbinom() functions:
    • Now have both definitions of the Negative Binomial (number of trials or number of failures).
  • Replaced Shiny module Distributacalcul_vis() by distributacalculVis() :
    • The new visualisation function can choose which modules to execute.
    • There are 4 different modules for :
      1. The input parameters parametersBox().
      2. The PDF and CDF functionsBox().
      3. The risk measures (VaR and TVaR) riskMeasuresBox().
      4. Various moments momentsBox().
    • Added LaTeX formulas for all functions (VaR and TVaR, moments, PDF and CDF).
    • Added language translation with package shiny.i18n.
      • English and French translation.
      • Added shiny.i18n to the imports section of the DESCRIPTION file.
      • Changed titles and user inputs to be bilingual.
      • Added (bilingual) labels to the plots.
      • Added language input at the top right of the screen.
      • Added a json file containing the translations under man-roxygen/translations.
    • Removed old function.
  • Updated Shiny functions with new syntax.
  • Updated vignette with new syntax.
  • Updated README with new syntax.
  • Added structure for tests.
    • Included basic tests to learn how it works.
  • Added function to generate plot of simulation observations from a copula.

Bug Fixes and Changes

  • Added pPareto() and dPareto() functions.
    • Fixed environment error in the quantile function plot of the Shiny app.
  • Changed the name of file lowerthan-template to lower.tail-template :
    • Typo, it contains the template for the lower.tail parameter.
    • Updated binomial distribution template file reference to the new one.
  • Fixed the documentation of the compound distributions.
    • Now one input under BINCOMP / BNCOMP / PCOMP for the family of distributions.
    • All functions appear there instead of having their individual entries in the documentation.
  • Improved title case and style according to the Tidyverse style guide.
  • Changed filename of riskMeasuresBoxVis to riskMeasuresBox
  • Fixed mistake in discrete uniform template file which linked the PDF instead of the CDF in the return field of the description.
  • Added rmarkdown to the vignette builder section of the DESCRIPTION file as per this rubric of the writing R extensions guide.
  • Added CITATION file for custom citation.
  • Converted parameters into templates for both Shiny applications.
  • Added more translations for the second Shiny app.

Minor Changes

  • Changed Etronq_ functions to Etrunc_ functions.
    • French spelling to English spelling.
    • Later on, with change of all syntax, changed to expValTrunc
  • Changed Pareto PDF and CDF from p_ and d_ to p and d:
    • To be consistent with other packages.
    • Also updated return field of the pareto-template file.
  • Optimized compound distributions’ p_ (CDF) and TVaR_ codes
    • Removed sapply and used R’s vector operations instead
  • Merged documentation of families under one file.
    • The procedure is to :
      1. Create a new file under the distribution’s name
      2. Copy all functions to it.
      3. Remove all individual function files.
      4. Remove the template file from man-roxygen/.
    • This helps simplify the package and the documentation.
    • The same procedure is used by other stats packages.
    • Distributions merged :
      1. Binomial (06/16/2020).
      2. Compound distributions (06/16/2020).
      3. All of them (08/17/2020)
  • Replaced(ing) @templateVar calls for @template calls with files for each variable.
    • Makes it so each distribution’s template file doesn’t individually define the parameters and it’s easier to change the definition of a parameter for all the distributions.
    • Removed(ing) the conditional calls for the variables from the distributions’ template files.
    • Done: beta, burr, uniform (both), exp, gamma, hyper, IG.
  • Pulled x, q, min and max, and lower.tail parameter descriptions directly from functions of the stats package with @inheritParams call.
  • Added notes for VatR functions which are just wrappers and added direct links to the underlying functions.
  • Removed special symbols in LaTeX formulas.
  • Added distributacalcul-package.R file which describes the package !
  • Added author, and did other minor changes, to the vignette.
  • Added %>% from dplyr to NAMESPACE file.
  • Fixed documentation of Copula functions.

Distributacalcul 0.2.2 2020-06-13

New Features

  • Added Pareto PDF and CDF.
  • Updated IG VaR from ‘statmod’ package.

Bug Fixes

  • Updated URL in description file to GitHub.io site.

Distributacalcul 0.2.1 Unreleased

Unremarkable release which updates the README.

Bug Fixes

  • Updated table of functions.
  • Fixed English grammar mistake in README example.
  • Updated instructions for downloading the package.

Distributacalcul 0.2.0 2020-06-09

New Features

  • Added the expected value and variance of the Poisson distribution as functions for eventual addition to Shiny application.
  • Added the MGF of the Erlang distribution to ensure all distributions which have a MGF have the function defined.
  • Added NEWS.md file.

Bug Fixes

  • Added @return field to function documentation to describe what each functions returns.
  • Changed the filenames for d_negbinom and p_negbinom to dnegbinom and pnegbinom as per function name.
  • Changed function names for E_logarithmique and V_logarithmique to E_logarithmic and V_logarithmic as per other english function names.
  • Changed all ‘F’ and ‘T’ to ‘FALSE’ and ‘TRUE’.
  • Fixed PGF to MGF in description of MGF_pois function.
  • Removed ‘dontrun’ from the TVaR_BNCOMP example.
  • Removed d_negbinom and p_negbinom (instead of not exporting the function) until I can fix the mistake in the function.
  • Removed the notes on Distributacalcul_vis module functions and replaced with ‘return’ field describing their use.
  • Added domain restrictions for the Poisson distribution.

Distributacalcul 0.1.0 Unreleased

  • initial release.
  • Fixed original CRAN comment on license by changing the file license to be MIT.