ldt: Let Data Talk

Methods and tools for creating a model set and estimating and evaluating the explanation or prediction power of its members. 'SUR' modelling (for parameter estimation), 'logit'/'probit' modelling (for binary classification), and 'VARMA' modelling (for time-series forecasting) are implemented. Evaluations are both in-sample and out-of-sample. It can be used for stepwise regression analysis <https://en.wikipedia.org/wiki/Stepwise_regression>, automatic model selection and model averaging (Claeskens and Hjort (2008, ISBN:1139471805, 9781139471800)), calculating benchmarks, and doing sensitivity analysis (Leamer (1983) <https://www.jstor.org/stable/1803924> proposal).

Version: 0.1.1.0
Depends: R (≥ 3.5.0)
Imports: Rcpp, readxl, jsonlite
LinkingTo: BH, Rcpp
Suggests: knitr, testthat, rmarkdown, kableExtra, MASS
Published: 2023-01-16
Author: Ramin Mojab [aut, cre], Microsoft Corporation [cph] (MIT license. Code from LightGBM package is used for AUC calculations.), Michael Hutt [cph] (MIT license. Original code for Nelder-Mead algorithm.), Stephen Becker [cph] (BSD 3-clause license. Original code for Nelder-Mead algorithm. The L-BFGS-B algorithm was written in the 1990s (mainly 1994, some revisions 1996) by Ciyou Zhu (in collaboration with R.H. Byrd, P. Lu-Chen and J. Nocedal). L-BFGS-B Version 3.0 is an algorithmic update from 2011, with coding changes by J. L. Morales), Math.NET [cph] (MIT license. Code from the 'Math.NET Numerics' library is used in calculating running statistics.), Christian Ammer [cph] (CC BY-SA 3.0 license. Code is used for transosing a matrix.)
Maintainer: Ramin Mojab <rmojab63 at gmail.com>
License: GPL (≥ 3)
Copyright: see file COPYRIGHTS
URL: https://github.com/rmojab63/LDT
NeedsCompilation: yes
SystemRequirements: C++17
CRAN checks: ldt results

Documentation:

Reference manual: ldt.pdf
Vignettes: DC: Loan default prediction
DC: Credit card fraud detection
SUR: Determinants of long-run economic growth
SUR: A simulation
VARMA: Predictability of the commodity prices

Downloads:

Package source: ldt_0.1.1.0.tar.gz
Windows binaries: r-devel: not available, r-release: ldt_0.1.1.0.zip, r-oldrel: not available
macOS binaries: r-release (arm64): not available, r-oldrel (arm64): not available, r-release (x86_64): not available, r-oldrel (x86_64): not available

Linking:

Please use the canonical form https://CRAN.R-project.org/package=ldt to link to this page.