modeltime 1.1.0 Unreleased

Spark Backend

New Algorithms: Smooth Package Integration

If users install smooth, the following models become available:

  • adam_reg(): Interfaces with the ADAM forecasting algorithm in smooth.

  • exp_smoothing(): A new engine “smooth_es” connects to the Exponential Smoothing algorithm in smooth::es(). This algorithm has several advantages, most importantly that it can use x-regs (unlike “ets” engine).

Nested Modeltime Improvements

(potentially) Breaking Changes

  • extract_nested_train_split and extract_nested_test_split: Changed parameter from .data to .object for consistency with other “extract” functions

  • Added a new logged feature to modeltime_nested_fit() to track the attribute “metric_set”, which is needed for ensembles. Old nested modeltime objects will need to be re-run to get this new attribute. This will be used in ensembles.

modeltime 1.0.0 2021-09-14

New Feature: Nested (Iterative) Forecasting

Nested (Iterative) Forecasting is aimed at making it easier to perform forecasting that is traditionally done in a for-loop with models like ARIMA, Prophet, and Exponential Smoothing. Functionality has been added to:

Format data in a Nested Time Series structure

Nested Model Fitting (Train/Test)

Nested Model Selection

Nested Model Refitting (Actual Data)

New Vignette

Vignette Improvements

  • Forecasting with Global Models: Added more complete steps in the forecasting process so now user can see how to forecast each step from start to finish including future forecasting.

New Accuracy Metric Set and Yardstick Functions

  • extended_forecast_accuracy_metric_set(): Adds the new MAAPE metric for handling intermittent data when MAPE returns Inf.
  • maape(): New yardstick metric that calculates “Mean Arctangent Absolute Percentage Error” (MAAPE). Used when MAPE returns Inf typically due to intermittent data.


modeltime 0.7.0 2021-07-16

Group-Wise Accuracy and Confidence Interval by Time Series ID

We’ve expanded Panel Data functionality to produce model accuracy and confidence interval estimates by a Time Series ID (#114). This is useful when you have a Global Model that produces forecasts for more than one time series. You can more easily obtain grouped accuracy and confidence interval estimates.

  • modeltime_calibrate(): Gains an id argument that is a quoted column name. This identifies that the residuals should be tracked by an time series identifier feature that indicates the time series groups.

  • modeltime_accuracy(): Gains a acc_by_id argument that is TRUE/FALSE. If the data has been calibrated with id, then the user can return local model accuracy by the identifier column. The accuracy data frame will return a row for each combination of Model ID and Time Series ID.

  • modeltime_forecast(): Gains a conf_by_id argument that is TRUE/FALSE. If the data has been calibrated with id, then the user can return local model confidence by the identifier column. The forecast data frame will return an extra column indicating the identifier column. The confidence intervals will be adjusted based on the local time series ID variance instead of the global model variance.

New Vignette

Forecasting Panel Data

New Algorithms

THIEF: Temporal Hierarchical Forecasting

  • temporal_hierarchy(): Implements the thief package by Rob Hyndman and Nikolaos Kourentzes for “Temporal HIErarchical Forecasting”. #117

Bug Fixes

modeltime 0.6.1 2021-06-13

Parallel Processing

Bug Fixes

  • Issue #110: Fix bug with cores > cores_available.

modeltime 0.6.0 2021-05-30

Workflowset Integration

modeltime_fit_workflowset() (#85) makes it easy to convert workflow_set objects to Modeltime Tables (mdl_time_tbl). Requires a refitting process that can now be performed in parallel or in sequence.

New Algorithms

New Dials Parameters

exp_smoothing() gained 3 new tunable parameters:

  • smooth_level(): This is often called the “alpha” parameter used as the base level smoothing factor for exponential smoothing models.
  • smooth_trend(): This is often called the “beta” parameter used as the trend smoothing factor for exponential smoothing models.
  • smooth_seasonal(): This is often called the “gamma” parameter used as the seasonal smoothing factor for exponential smoothing models.

Parallel Processing

Updates for parsnip >= 0.1.6

General Improvements

  • Improve Model Description of Recursive Models (#96)

Potential Breaking Changes

  • We’ve added new parameters to Exponential Smoothing Models. exp_smoothing() models produced in prior versions may require refitting with modeltime_refit() to upgrade their internals with the new parameters.

modeltime 0.5.1 2021-04-03

Recursive Ensemble Predictions

  • Add support for recursive() for ensembles. The new recursive ensemble functionality is in modeltime.ensemble >=

modeltime 0.5.0 2021-03-29

Recursive Panel Predictions

Breaking Changes

modeltime 0.4.2 2021-03-19

New Algorithms

Baseline algorithms (#5, #37) have been created for comparing high-performance methods with simple forecasting methods.

  • window_reg: Window-based methods such as mean, median, and even more complex seasonal models based on a forecasting window. The main tuning parameter is window_size.
  • naive_reg: NAIVE and Seasonal NAIVE (SNAIVE) Regression Models

Yardstick Helpers

  • metric_tweak() - Can modify yardstick metrics like mase(), which have seasonal parameters.
  • default_forecast_accuracy_metric_set() - Gets a ... parameter that allows us to add more metrics beyond the defaults.

Modeltime Residual Tests

A new function is added modeltime_residuals_test() (#62, #68). Tests are implemented:

  • Shapiro Test - Test for Normality of residuals
  • Box-Pierce, Ljung-Box, and Durbin-Watson Tests - Test for Autocorrelation of residuals


modeltime 0.4.1 2021-01-17


modeltime 0.4.0 2020-11-23

New Functions

Breaking Changes

  • Removed arima_workflow_tuned dataset.

modeltime 0.3.1 2020-11-09

as_modeltime_table(): New function to convert one or more fitted models stored in a list to a Modeltime Table.

Bug Fixes

  • Update m750_models: Fixes error “R parsnip Error: Internal error: Unknown composition type.”

modeltime 0.3.0 2020-10-28

Panel Data

modeltime_forecast() upgrades:

  • keep_data: Gains a new argument keep_data. This is useful when the new_data and actual_data has important information needed in analyzing the forecast.
  • arrange_index: Gains a new argument arrange_index. By default, the forecast keeps the rows in the same order as the incoming data. Prior versions arranged Model Predictions by .index, which impacts the users ability to match to Panel Data which is not likely to be arranged by date. Prediction best-practices are to keep the original order of the data, which will be preserved by default. To get the old behavior, simply toggle arrange_index = TRUE.

modeltime_calibrate(): Can now handle panel data.

modeltime_accuracy(): Can now handle panel data.

plot_modeltime_forecast(): Can handle panel data provided the data is grouped by an ID column prior to plotting.

Error Messaging


  • Compatibility with parsnip >= 0.1.4. Uses set_encodings() new parameter allow_sparse_x.

modeltime 0.2.1 2020-10-08


  • modeltime_refit() - Changes to improve fault tolerance and error handling / messaging when making ensembles.

modeltime 0.2.0 2020-09-28


  • Integrates modeltime.ensemble, a new R package designed for forecasting with ensemble models.

New Workflow Helper Functions


  • Documentation - Algorithms now identify default parameter values in the “Engine Details” Section in their respective documentation. E.g. ?prophet_boost
  • prophet_reg() can now have regressors controlled via set_engine() using the following parameters:
    • regressors.mode - Set to seasonality.mode by default.
    • regressors.prior.scale - Set to 10,000 by default.
    • regressors.standardize - Set to “auto” by default.

Data Sets

Modeltime now includes 4 new data sets:

  • m750 - M750 Time Series Dataset
  • m750_models - 3 Modeltime Models made on the M750 Dataset
  • m750_splits - An rsplit object containing Train/test splits of the M750 data
  • m750_training_resamples - A Time Series Cross Validation time_series_cv object made from the training(m750_splits)

Bug Fix

modeltime 0.1.0 2020-09-02

New Features

Forecast without Calibration/Refitting

Sometimes it’s important to make fast forecasts without calculating out-of-sample accuracy and refitting (which requires 2 rounds of model training). You can now bypass the modeltime_calibrate() and modeltime_refit() steps and jump straight into forecasting the future. Here’s an example with h = "3 years". Note that you will not get confidence intervals with this approach because calibration data is needed for this.

# Make forecasts without calibration/refitting (No Confidence Intervals)
# - This assumes the models have been trained on m750
) %>%
        h = "3 years",
        actual_data = m750
    ) %>%
    plot_modeltime_forecast(.conf_interval_show = F)

Residual Analysis & Diagonstics

A common tool when forecasting and analyzing residuals, where residuals are .resid = .actual - .prediction. The residuals may have autocorrelation or nonzero mean, which can indicate model improvement opportunities. In addition, users may which to inspect in-sample and out-of-sample residuals, which can display different results.

New Models


Use seasonal_reg() and set engine to “tbats”.

    seasonal_period_1 = "1 day",
    seasonal_period_2 = "1 week"
) %>% 


Use nnetar_reg() and set engine to “nnetar”.

model_fit_nnetar <- nnetar_reg() %>%

Prophet Model - Logistic Growth Support

  • prophet_reg() and prophet_boost():
    • Now supports logistic growth. Set growth = 'logistic' and one or more of logistic_cap and logistic_floor to valid saturation boundaries.
    • New arguments making it easier to modify the changepoint_num, changepoint_range, seasonality_yearly, seasonality_weekly, seasonality_daily, logistic_cap, logistic_floor

New Workflow Helper Functions


  • modeltime_refit(): When modeltime model parameters update (e.g. when Auto ARIMA changes to a new model), the Model Description now alerts the user (e.g. “UPDATE: ARIMA(0,1,1)(1,1,1)[12]”).

  • modeltime_calibrate(): When training data is supplied in a time window that the model has previously been trained on (e.g. training(splits)), the calibration calculation first inspects whether the “Fitted” data exists. If it iexists, it returns the “Fitted” data. This helps prevent sequence-based (e.g. ARIMA, ETS, TBATS models) from displaying odd results because these algorithms can only predict sequences directly following the training window. If “Fitted” data is being used, the .type column will display “Fitted” instead of “Test”.

Bug Fixes

  • modeltime_forecast():

    • Implement actual_data reconciliation strategies when recipe removes rows. Strategy attempts to fill predictors using “downup” strategy to prevent NA values from removing rows.
    • More descriptive errors when external regressors are required.
  • modeltime_accuracy(): Fix issue with new_data not recalibrating.

  • prophet_reg() and prophet_boost() - Can now perform logistic growth growth = 'logistic'. The user can supply “saturation” bounds using logistic_cap and/or logisitc_floor.

Breaking Changes

modeltime 0.0.2 2020-07-03

Confidence Interval Estimation

  • modeltime_forecast(): Now estimates confidence intervals using centered standard deviation. The mean is assumed to be zero and residuals deviate from mean = 0.


  • Updates to work with parsnip 0.1.2.
  • prophet_boost(): Set nthreads = 1 (default) to ensure parallelization is thread safe.

modeltime 0.0.1 2020-06-22

  • Initial Release