Function to choose optimal forecast model for monthly data

MonthModelPicker(data, Outcome, DateVar, H.Horizon = 14)

Arguments

data	A monthly tsibble.
Outcome	A valid variable name for the Outcome to be modelled in in `data`.
DateVar	A valid variable name for the time index in `data`.
H.Horizon	An integer for the forecast horizon/test subset of `data`.

Value

A list containing:

Accuracy.Table: accuracy for the forecast horizon against the test sample.
test: the test set
train: the training set
Model.Fits: the model fits
Model.Forecasts: the forecasts
Min.Model: the minimum model by MAE
Min.Report: the minimum model report
Min.Res.Plot: the gg_tsdisplay for the minimum model
Min.Forecast.Plot: a plot of the minimum forecast by MAE

Examples

data(NWSM)
MonthModelPicker(NWSM, MaxAvg, Month, H.Horizon=12)
#> Warning: Current temporal ordering may yield unexpected results.
#> ℹ Suggest to sort by ``, `Month` first.
#> Joining, by = c("MO", "YR", "MaxAvg", "MinAvg", "Precip", "Snow", "Month")
#> Warning: Current temporal ordering may yield unexpected results.
#> ℹ Suggest to sort by ``, `Month` first.
#> Warning: Current temporal ordering may yield unexpected results.
#> ℹ Suggest to sort by ``, `Month` first.
#> Warning: Current temporal ordering may yield unexpected results.
#> ℹ Suggest to sort by ``, `Month` first.
#> Series: MaxAvg 
#> Model: COMBINATION 
#> Combination: (MaxAvg + MaxAvg) * 0.5
#> 
#> ====================================
#> 
#> Series: MaxAvg + MaxAvg 
#> Model: COMBINATION 
#> Combination: MaxAvg + MaxAvg
#> 
#> ============================
#> 
#> Series: MaxAvg 
#> Model: ARIMA(1,0,0)(1,1,0)[12] w/ drift 
#> 
#> Coefficients:
#>          ar1     sar1  constant
#>       0.2471  -0.4665    0.1588
#> s.e.  0.0663   0.0618    0.2291
#> 
#> sigma^2 estimated as 11.12:  log likelihood=-566.58
#> AIC=1141.16   AICc=1141.35   BIC=1154.66
#> 
#> Series: MaxAvg 
#> Model: ETS(A,N,A) 
#>   Smoothing parameters:
#>     alpha = 0.1817322 
#>     gamma = 0.0001008462 
#> 
#>   Initial states:
#>         l        s1        s2        s3       s4       s5       s6       s7
#>  63.30714 -17.44621 -10.46094 0.6837303 12.23725 18.48901 18.54583 11.03021
#>        s8        s9       s10       s11       s12
#>  5.147654 -1.775394 -7.626994 -12.27733 -16.54682
#> 
#>   sigma^2:  7.0556
#> 
#>      AIC     AICc      BIC 
#> 1698.915 1701.179 1750.355 
#> 
#> 
#> $test
#> # A tsibble: 12 x 7 [1M]
#>       MO    YR MaxAvg MinAvg Precip  Snow    Month
#>    <int> <int>  <dbl>  <dbl>  <dbl> <dbl>    <mth>
#>  1    12  2019   48.1   38     4.39   0   2019 Dec
#>  2    11  2019   54.4   39.2   1.52   0   2019 Nov
#>  3    10  2019   60.8   42.1   1.51   0   2019 Oct
#>  4     9  2019   72.1   56.8   3.85   0   2019 Sep
#>  5     8  2019   82.6   60.9   1.23   0   2019 Aug
#>  6     7  2019   80.3   60.3   0.8    0   2019 Jul
#>  7     6  2019   76.6   54.8   0.45   0   2019 Jun
#>  8     5  2019   72.7   51.6   1.51   0   2019 May
#>  9     4  2019   62.6   46.1   2.98   0   2019 Apr
#> 10     3  2019   56.6   36.7   1.54   0   2019 Mar
#> 11     2  2019   42.9   32.2   4.1    6.5 2019 Feb
#> 12     1  2019   49.2   37.1   2.79   0   2019 Jan
#> 
#> $train
#> # A tsibble: 228 x 7 [1M]
#>       MO    YR MaxAvg MinAvg Precip  Snow    Month
#>    <int> <int>  <dbl>  <dbl>  <dbl> <dbl>    <mth>
#>  1     1  2000   45.1   34.5   5.66     1 2000 Jan
#>  2     2  2000   50.2   37.1   4.5      0 2000 Feb
#>  3     3  2000   53.9   37.3   3.21     0 2000 Mar
#>  4     4  2000   64.1   43.9   1.82     0 2000 Apr
#>  5     5  2000   66.0   49.1   2.7      0 2000 May
#>  6     6  2000   76.4   54.1   1.19     0 2000 Jun
#>  7     7  2000   78.4   57.2   0.15     0 2000 Jul
#>  8     8  2000   78.6   56.9   0.12     0 2000 Aug
#>  9     9  2000   73.9   53.9   1.67     0 2000 Sep
#> 10    10  2000   63.2   47.2   3.25     0 2000 Oct
#> # … with 218 more rows
#> 
#> $Model.Fits
#> # A mable: 1 x 10
#>                       `K = 1`                     `K = 2`
#>                       <model>                     <model>
#> 1 <LM w/ ARIMA(0,1,5) errors> <LM w/ ARIMA(0,1,1) errors>
#> # … with 8 more variables: K = 3 <model>, ARIMA <model>, ETS <model>,
#> #   NNET1 <model>, NNET2 <model>, NNET3 <model>, prophet <model>,
#> #   Combo1 <model>
#> 
#> $Model.Forecasts
#> # A fable: 120 x 4 [1M]
#> # Key:     .model [10]
#>    .model    Month    MaxAvg .mean
#>    <chr>     <mth>    <dist> <dbl>
#>  1 K = 1  2019 Jan N(48, 11)  48.1
#>  2 K = 1  2019 Feb N(51, 12)  50.6
#>  3 K = 1  2019 Mar N(57, 12)  56.9
#>  4 K = 1  2019 Apr N(65, 12)  64.8
#>  5 K = 1  2019 May N(73, 12)  73.2
#>  6 K = 1  2019 Jun N(80, 13)  79.9
#>  7 K = 1  2019 Jul N(83, 13)  82.9
#>  8 K = 1  2019 Aug N(81, 14)  81.1
#>  9 K = 1  2019 Sep N(75, 15)  75.1
#> 10 K = 1  2019 Oct N(67, 15)  66.6
#> # … with 110 more rows
#> 
#> $Accuracy.Table
#> # A tibble: 10 x 10
#>    .model  .type    ME  RMSE   MAE   MPE  MAPE  MASE RMSSE    ACF1
#>    <chr>   <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>   <dbl>
#>  1 ARIMA   Test  -1.32  3.20  2.55 -2.01  4.21   NaN   NaN  0.0525
#>  2 Combo1  Test  -1.78  3.45  2.37 -3.02  4.02   NaN   NaN -0.0668
#>  3 ETS     Test  -2.23  3.97  2.64 -4.02  4.62   NaN   NaN -0.0806
#>  4 K = 1   Test  -2.41  3.50  2.85 -4.32  5.00   NaN   NaN -0.357 
#>  5 K = 2   Test  -2.17  3.93  2.75 -3.91  4.77   NaN   NaN -0.0567
#>  6 K = 3   Test  -2.18  3.93  2.73 -3.92  4.73   NaN   NaN -0.0556
#>  7 NNET1   Test  -1.90  4.27  3.21 -3.50  5.78   NaN   NaN -0.0426
#>  8 NNET2   Test  -1.82  4.69  3.53 -3.46  6.42   NaN   NaN -0.0101
#>  9 NNET3   Test  -1.83  5.12  3.70 -3.51  6.76   NaN   NaN -0.0639
#> 10 prophet Test  -2.28  4.30  3.03 -4.09  5.24   NaN   NaN -0.0119
#> 
#> $Min.Model
#> # A tibble: 1 x 10
#>   .model .type    ME  RMSE   MAE   MPE  MAPE  MASE RMSSE    ACF1
#>   <chr>  <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>   <dbl>
#> 1 Combo1 Test  -1.78  3.45  2.37 -3.02  4.02   NaN   NaN -0.0668
#> 
#> $Min.Report
#> # A mable: 1 x 1
#>          Combo1
#>         <model>
#> 1 <COMBINATION>
#> 
#> $Min.Res.Plot
#>