Package index • tscv

Time Series Cross-Validation

Functions and utilities for time series cross-validation

make_accuracy(): Estimate accuracy metrics to evaluate point forecast

make_errors(): Calculate forecast errors and percentage errors

make_future(): Convert the forecasts from a fable to a future_frame

make_split(): Create a split_frame for train and test splits per time series.

make_tsibble(): Convert tibble to tsibble

slice_train(): Slice the train data from the complete data

slice_test(): Slice the test data from the complete data

Data visualization

Functions and utilities for data visualization

plot_bar(): Plot data as bar chart

plot_density(): Plot the density via Kernel Density Estimator

plot_histogram(): Plot data as histogram

plot_line(): Plot data as line chart

plot_point(): Plot data as scatterplot

plot_qq(): Quantile-Quantile plot

theme_tscv(): Custom ggplot2 theme for tscv package

theme_tscv_dark(): Dark ggplot2 theme for tscv package

scale_color_tscv(): Color scale constructor for tscv colors.

scale_fill_tscv(): Fill scale constructor for tscv colors.

tscv_cols(): Function to extract tscv colors as hex codes.

tscv_pal(): Return function to interpolate a tscv color palette.

Forecasting

DSHW

The DSHW model and its supported methods

DSHW(): Automatic train a DSHW model

forecast(<DSHW>): Forecast a trained DSHW model

fitted(<DSHW>): Extract fitted values from a trained DSHW model

residuals(<DSHW>): Extract residuals from a trained DSHW model

model_sum(<DSHW>): Provide a succinct summary of a trained DSHW model

train_dshw(): Double Seasonal Holt-Winters model

ELM

The ELM model and its supported methods

ELM(): Extreme Learning Machine (ELM)

forecast(<ELM>): Forecast a trained ELM model

fitted(<ELM>): Extract fitted values from a trained ELM model

residuals(<ELM>): Extract residuals from a trained ELM model

model_sum(<ELM>): Provide a succinct summary of a trained ELM model

train_elm(): Extreme Learning Machine (ELM)

MLP

The MLP model and its supported methods

MLP(): Automatic training of MLPs

forecast(<MLP>): Forecast a trained MLP model

fitted(<MLP>): Extract fitted values from a trained MLP model

residuals(<MLP>): Extract residuals from a trained MLP model

model_sum(<MLP>): Provide a succinct summary of a trained MLP model

train_mlp(): Multilayer Perceptron (MLP)

SMEAN

The SMEAN model and its supported methods

SMEAN(): Seasonal mean model

forecast(<SMEAN>): Forecast a trained seasonal mean model

fitted(<SMEAN>): Extract fitted values from a trained seasonal mean model

residuals(<SMEAN>): Extract residuals from a trained seasonal mean model

model_sum(<SMEAN>): Provide a succinct summary of a trained seasonal mean model

train_smean(): Seasonal mean model

SMEDIAN

The SMEDIAN model and its supported methods

SMEDIAN(): Seasonal median model

forecast(<SMEDIAN>): Forecast a trained seasonal median model

fitted(<SMEDIAN>): Extract fitted values from a trained seasonal median model

residuals(<SMEDIAN>): Extract residuals from a trained seasonal median model

model_sum(<SMEDIAN>): Provide a succinct summary of a trained seasonal median model

train_smedian(): Seasonal median model

MEDIAN

The MEDIAN model and its supported methods

MEDIAN(): Median model

forecast(<MEDIAN>): Forecast a trained median model

fitted(<MEDIAN>): Extract fitted values from a trained median model

residuals(<MEDIAN>): Extract residuals from a trained median model

model_sum(<MEDIAN>): Provide a succinct summary of a trained median model

train_median(): Median model

SNAIVE2

The SNAIVE2 model and its supported methods

SNAIVE2(): Seasonal naive model

forecast(<SNAIVE2>): Forecast a trained seasonal naive model

fitted(<SNAIVE2>): Extract fitted values from a trained seasonal naive model

residuals(<SNAIVE2>): Extract residuals from a trained seasonal naive model

model_sum(<SNAIVE2>): Provide a succinct summary of a trained seasonal naive model

train_snaive2(): Seasonal naive model

TBATS

The TBATS model and its supported methods

TBATS(): Automatic train a TBATS model

forecast(<TBATS>): Forecast a trained TBATS model

fitted(<TBATS>): Extract fitted values from a trained TBATS model

residuals(<TBATS>): Extract residuals from a trained TBATS model

model_sum(<TBATS>): Provide a succinct summary of a trained TBATS model

train_tbats(): TBATS model

EXPERT

The EXPERT model and its supported methods

EXPERT(): Automatic train a EXPERT model

forecast(<EXPERT>): Forecast a trained EXPERT model

fitted(<EXPERT>): Extract fitted values from a trained EXPERT model

residuals(<EXPERT>): Extract residuals from a trained EXPERT model

model_sum(<EXPERT>): Provide a succinct summary of a trained EXPERT model

train_expert(): EXPERT model

Miscellaneous

Functions and utilities for data preparation, etc.

estimate_mode(): Estimate mode of a distribution based on Kernel Density Estimation

estimate_kurtosis(): Estimate kurtosis

estimate_skewness(): Estimate skewness

acf_vec(): Estimate the sample autocorrelation of a numeric vector

estimate_acf(): Estimate the sample autocorrelation

pacf_vec(): Estimate the sample partial autocorrelation of a numeric vector

estimate_pacf(): Estimate the sample partial autocorrelation

interpolate_missing(): Interpolate missing values

smooth_outlier(): Identify and replace outliers

lst_to_env(): Assign objects within a list to an environment

check_data(): Check, convert and shape the input data

summarise_data(): Summary statistics for time series data

summarise_stats(): Summary statistics for time series data

summarise_split(): Summary table of the splitting into training and testing

initialize_split(): Initialize a plan for train-test split

split_index(): Create indices for train and test splits.

expand_split(): Expand the split_frame

file_name(): Create a name for a folder or file

number_string(): Helper function to create numbered strings.

`%out%`: Negated value matching

glue_header(): Create a header from text

log_platform(): Create platform info

log_time(): Create string with elapsed time.

Data sets

Example data sets

elec_load: Hourly electricity load (actual values and forecasts)

elec_price: Hourly day-ahead electricity spot prices

M4_monthly_data: Monthly time series data from the M4 Competition

M4_quarterly_data: Quarterly time series data from the M4 Competition