[R] Extracting the MAPE value from a fitted Time Series Model

[Paul Bernal]

Dear friends,

I want to extract the MAPE value from a fitted time series model. This is
what I have:

> str(TransitSpline)
List of 12
 $ method               : chr "Cubic Smoothing Spline"
 $ level                : num [1:2] 80 95
 $ x                    : Time-Series [1:385] from 1 to 385: 77 75 85 74 73
96 82 90 91 81 ...
 $ series               : chr "data$Transits"
 $ mean                 : Time-Series [1:10, 1] from 386 to 395: 186 178
170 163 155 ...
  ..- attr(*, "dimnames")=List of 2
  .. ..$ : NULL
  .. ..$ : chr "Series 1"
 $ upper                : Time-Series [1:10, 1:2] from 386 to 395: 202 199
197 197 197 ...
  ..- attr(*, "dimnames")=List of 2
  .. ..$ : NULL
  .. ..$ : chr [1:2] "Series 1" "Series 2"
 $ lower                : Time-Series [1:10, 1:2] from 386 to 395: 171 158
144 129 113 ...
  ..- attr(*, "dimnames")=List of 2
  .. ..$ : NULL
  .. ..$ : chr [1:2] "Series 1" "Series 2"
 $ model                :List of 2
  ..$ beta: num 6.15
  ..$ call: language splinef(y = data$Transits)
 $ fitted               : Time-Series [1:385] from 1 to 385: 76.1 77.3 78.5
80.1 82.2 ...
 $ residuals            : Time-Series [1:385] from 1 to 385: NA -1.3 9.49
-8.64 -4.34 ...
 $ standardizedresiduals: Time-Series [1:385] from 1 to 385: NA -0.875
6.517 -5.586 -2.736 ...
 $ onestepf             : Time-Series [1:385] from 1 to 385: NA 76.3 75.5
82.6 77.3 ...
 - attr(*, "class")= chr [1:2] "splineforecast" "forecast"


> str(summary(TransitSpline))
#Here I want to get the value for the MAPE measure
Forecast method: Cubic Smoothing Spline

Model Information:
$`beta`
[1] 6.149167

$call
splinef(y = data$Transits)


Error measures:
                      ME     RMSE      MAE        MPE     MAPE     MASE
   ACF1
Training set -0.07776434 12.10204 9.003675 -0.2408687 5.377131 0.930913
-0.2766975

Forecasts:
    Point Forecast     Lo 80    Hi 80      Lo 95    Hi 95
386       186.0153 170.52426 201.5064 162.323777 209.7069
387       178.2220 157.87687 198.5671 147.106804 209.3372
388       170.4287 143.80863 197.0487 129.716832 211.1405
389       162.6353 128.61257 196.6581 110.602006 214.6687
390       154.8420 112.52646 197.1576  90.125956 219.5581
391       147.0487  95.66491 198.4324  68.463984 225.6334
392       139.2553  78.10706 200.4036  45.737114 232.7736
393       131.4620  59.92462 202.9994  22.055013 240.8690
394       123.6687  41.14798 206.1894  -2.535833 249.8732
395       115.8753  21.82457 209.9261 -27.962900 259.7136
'data.frame':   10 obs. of  5 variables:
 $ Point Forecast: num  186 178 170 163 155 ...
 $ Lo 80         : num  171 158 144 129 113 ...
 $ Hi 80         : num  202 199 197 197 197 ...
 $ Lo 95         : num  162.3 147.1 129.7 110.6 90.1 ...
 $ Hi 95         : num  210 209 211 215 220 ...

any idea on how to accomplish this?

Best regards,

Paul

	[[alternative HTML version deleted]]