dymo {dymo}

dymo

Description

Dynamic Mode Decomposition for Multivariate Time Feature Prediction

Usage

dymo(
  df,
  seq_len,
  n_windows = 10,
  ci = 0.8,
  smoother = FALSE,
  min_feats = NULL,
  max_feats = NULL,
  dates = NULL,
  error_scale = "naive",
  error_benchmark = "naive",
  seed = 42
)

Arguments

df

A data frame with time features on columns. You need at least two time features. In case of missing values, automatic missing imputation through kalman filter will be performed.

seq_len

Positive integer. Time-step number of the forecasting sequence. Default: NULL (automatic selection between 1 and the square root of full length).

n_windows

Positive integer. Number of validation windows to test prediction error. Default: 10.

ci

Confidence interval for prediction. Default: 0.8

smoother

Logical. Flag to TRUE for loess smoothing. Default: FALSE.

min_feats

Positive integer. Minimum number of time features to combine. Default: NULL (set equal to the total number of features)

max_feats

Positive integer. Maximum number of time features to combine. Default: NULL (set equal to the total number of features)

dates

Date. Vector with dates for time features.

error_scale

String. Scale for the scaled error metrics. Two options: "naive" (average of naive one-step absolute error for the historical series) or "deviation" (standard error of the historical series). Default: "naive".

error_benchmark

String. Benchmark for the relative error metrics. Two options: "naive" (sequential extension of last value) or "average" (mean value of true sequence). Default: "naive".

seed

Positive integer. Random seed. Default: 42.

Value

This function returns a list including:

• comb_metrics: error metrics for all possible combinations of time features (for each combination, pred_score, me, mae, mse, rmsse, mpe, mape, rmae, rrmse, rame, mase, smse, sce, gmrae, are averaged across features, ranks and validation windows)

• best_model: best combination resulting from the average prediction score across different ranks and features, including:

  • best_combination: combination of indexes and rank for the best model

  • testing_errors: testing errors for each time feature averaged across validation windows

  • quant_preds: min, max, q25, q50, q75, quantiles at selected ci, mean, sd, mode, skewness, kurtosis, IQR to range, median range ratio, upside probability and divergence for each point fo predicted sequences

  • plots: standard plot with confidence interval for each time feature

• time_log

Author(s)

Maintainer: Giancarlo Vercellino [e-mail] [copyright holder]

Giancarlo Vercellino [e-mail]

See Also

Useful links:

• https://rpubs.com/giancarlo_vercellino/dymo

Examples

dymo(time_features[,c(2, 3, 4)], seq_len = 10, dates = time_features$dates)