Skip to content

NEWS.md

Latest commit

 

History

History
105 lines (65 loc) · 3.08 KB

NEWS.md

File metadata and controls

105 lines (65 loc) · 3.08 KB

version 0.36.0

  • Add sensitivity analysis to ahead::dynrmf

version 0.32.0

  • Add contextridge2f for adding context to ridge2f forecasting.
  • Add stackeridge2f for time series stacked generalization

version 0.25.0

  • Implement ANY MODEL+GARCH(1, 1) forecasting for stocks (stochastic volatility models)
  • armagarchf doesn't use the bootstrap by default anymore

version 0.24.0

  • add fitted values in ridge2f

version 0.22.0

  • add explicit clustering to volatility forecasting in mlarch

version 0.18.0

  • Add ML-ARCH model, see vignettes for more examples

version 0.17.0

  • Add conformal prediction to ridge2f (with KDE, bootstrap, block bootstrap and sequential split conformal prediction)
  • Faster install, less imports

version 0.14.0

  • Add fit_func and predict_func for custom fitting and prediction functions of ahead::dynrmf (using caret Machine Learning).
  • Add forecasting combinations based on ForecastComb, adding Ridge and Elastic Net to the mix.

version 0.11.0

  • Include tests (90% coverage). After cloning, run:
install.packages("covr")
covr::report()

version 0.10.0

version 0.9.0

  • progress bars for bootstrap (independent, circular block, moving block)

version 0.8.0

  • empirical marginals for R-Vine copula simulation
  • risk-neutralize simulations

version 0.7.0

  • moving block bootstrap in ridge2f, basicf and loessf, in addition to circular block bootstrap from 0.6.2
  • adjust R-Vine copulas on residuals for ridge2f simulation
  • new plots for simulations see (new) vignettes
  • split conformal prediction intervals (very very experimental and basic right now, too conservative)
  • Depends and selective Imports (beneficial to Python and rpy2 for installation time?)
  • getsimulations extracts simulations from a given time series (from ridge2f and basicf)
  • getreturns extracts returns/log-returns from multivariate time series
  • splitts splits time series using a proportion of data

version 0.6.2

  • Add Block Bootstrap to ridge2f
  • Add external regressors to ridge2f
  • Add clustering to ridge2f
  • Add Block Bootstrap to loessf
  • Create new vignettes for ridge2f and loessf

version 0.6.1

  • Align version with Python's
  • Temporarily remove dependency with cclust

version 0.6.0

  • Include basic methods: mean forecast, median forecast, random walk forecast

version 0.5.0

  • add dropout regularization to ridge2f
  • parallel execution for type_pi == bootstrap in ridge2f (done in R /!, experimental)
  • preallocate matrices for type_forecast == recursive in ridge2f

version 0.4.2

  • new attributes mean, lower bound, upper bound forecast as numpy arrays

version 0.4.1

  • use get_frequency to get series frequency as a number
  • create a function get_tscv_indices for getting time series cross-validation indices