comb_GLMNET.RdComputes forecast combination weights using GLMNET Regression (OLS) regression.
comb_GLMNET(x, custom_error = NULL)An object of class 'foreccomb'. Contains training set (actual values + matrix of model forecasts) and optionally a test set.
Returns an object of class ForecastComb::foreccomb_res with the following components:
Returns the best-fit forecast combination method.
Returns the individual input models that were used for the forecast combinations.
Returns the combination weights obtained by applying the combination method to the training set.
Returns the intercept of the linear regression.
Returns the fitted values of the combination method for the training set.
Returns range of summary measures of the forecast accuracy for the training set.
Returns forecasts produced by the combination method for the test set. Only returned if input included a forecast matrix for the test set.
Returns range of summary measures of the forecast accuracy for the test set. Only returned if input included a forecast matrix and a vector of actual values for the test set.
Returns the data forwarded to the method.
The function integrates the GLMNET Regression forecast combination implementation of the ForecastCombinations package into ForecastComb.
The results are stored in an object of class 'ForecastComb::foreccomb_res', for which separate plot and summary functions are provided.
library(ForecastComb)
#> Registered S3 methods overwritten by 'ForecastComb':
#> method from
#> plot.foreccomb_res ahead
#> predict.foreccomb_res ahead
#> print.foreccomb_res_summary ahead
#> summary.foreccomb_res ahead
#>
#> Attaching package: ‘ForecastComb’
#> The following object is masked from ‘package:ahead’:
#>
#> comb_OLS
data(electricity)
print(head(electricity))
#> arima ets nnet dampedt dotm Actual
#> Jan 2007 36980.16 35692.31 37047.91 35540.66 36044.28 36420
#> Feb 2007 33587.29 33708.15 34523.56 33962.34 33821.69 32901
#> Mar 2007 36005.55 37366.20 36049.72 37317.91 37119.29 34595
#> Apr 2007 30925.25 30550.24 30721.91 30356.77 30350.95 29665
#> May 2007 30394.78 29167.64 29241.89 28766.40 28910.84 30154
#> Jun 2007 28938.14 29004.18 29211.91 29006.25 28229.28 28607
forecasting_methods <- colnames(electricity)[1:5]
train_obs <- electricity[1:84, "Actual"]
train_pred <- electricity[1:84, forecasting_methods]
test_obs <- electricity[85:123, "Actual"]
test_pred <- electricity[85:123, forecasting_methods]
data <- ForecastComb::foreccomb(train_obs, train_pred, test_obs, test_pred)
(obj <- ahead::comb_GLMNET(data))
#> $Method
#> [1] "GLMNET Regression Regression"
#>
#> $Models
#> [1] "arima" "ets" "nnet" "dampedt" "dotm"
#>
#> $Fitted
#> Time Series:
#> Start = 1
#> End = 84
#> Frequency = 1
#> [1] 35878.50 33644.27 36312.31 30258.31 28893.30 28392.59 28694.17 28434.62
#> [9] 30025.91 31556.28 34047.23 36407.60 35963.88 32998.01 35858.16 30097.28
#> [17] 29779.11 28522.30 28643.47 28707.67 29808.18 31791.58 34098.84 35714.49
#> [25] 35842.60 33195.45 34771.13 29914.63 28704.55 27724.39 28470.85 27868.45
#> [33] 28789.39 31455.03 31379.17 34098.66 34341.19 31758.03 32756.58 28978.77
#> [41] 27811.85 27104.06 27316.63 26865.51 27921.23 29981.37 31421.46 34178.69
#> [49] 35300.14 32291.19 33613.51 28637.18 27660.20 26684.56 27086.58 27163.08
#> [57] 28000.63 30273.40 31623.79 33433.12 34253.89 30869.39 32134.89 27652.63
#> [65] 27438.39 26946.58 26815.60 26670.87 27478.89 29286.78 31475.12 33636.87
#> [73] 33979.47 31359.82 31939.39 28394.24 28058.86 26678.05 26912.53 26670.41
#> [81] 27233.13 29445.80 31089.23 33539.09
#>
#> $Accuracy_Train
#> ME RMSE MAE MPE MAPE
#> Test set 3.551339e-12 950.6999 751.5739 0.007510127 2.417785
#>
#> $Input_Data
#> $Input_Data$Actual_Train
#> Time Series:
#> Start = 1
#> End = 84
#> Frequency = 1
#> [1] 36420 32901 34595 29665 30154 28607 29072 29018 29266 32955 35385 36311
#> [13] 36885 34077 34807 32379 29319 28025 29145 28814 29183 32311 33274 35355
#> [25] 36490 32544 32810 28334 28186 27302 27562 27323 27487 29978 31259 33734
#> [37] 36770 32347 33348 28815 28455 26761 27403 26743 27508 30261 32806 37107
#> [49] 35851 31347 32919 27378 27766 26948 26901 27234 27644 29307 30865 33056
#> [61] 33447 32359 31196 29083 28783 26558 27145 27012 27086 30168 31317 33327
#> [73] 34581 31645 33922 29393 27537 25813 26397 26189 27265 29126 31246 31838
#>
#> $Input_Data$Forecasts_Train
#> Time Series:
#> Start = 1
#> End = 84
#> Frequency = 1
#> arima ets nnet dampedt dotm
#> 1 36980.16 35692.31 37047.91 35540.66 36044.28
#> 2 33587.29 33708.15 34523.56 33962.34 33821.69
#> 3 36005.55 37366.20 36049.72 37317.91 37119.29
#> 4 30925.25 30550.24 30721.91 30356.77 30350.95
#> 5 30394.78 29167.64 29241.89 28766.40 28910.84
#> 6 28938.14 29004.18 29211.91 29006.25 28229.28
#> 7 29657.04 28528.66 29203.88 28814.83 28655.94
#> 8 29153.90 28315.55 29166.40 28434.89 28301.77
#> 9 29961.03 30160.74 30119.06 30349.56 30275.80
#> 10 31588.09 31766.03 31651.94 32071.41 31901.11
#> 11 33837.47 33822.40 35515.04 34113.82 34026.36
#> 12 34867.32 36794.86 36655.04 37026.80 37067.55
#> 13 35889.38 35607.80 36822.33 35729.63 36300.97
#> 14 32798.91 33587.61 32255.68 33768.59 33773.06
#> 15 34306.37 36737.83 35292.80 36758.11 36800.75
#> 16 29966.78 30370.67 29980.12 30557.80 30436.55
#> 17 30392.01 29190.21 30635.55 29636.53 29688.95
#> 18 29015.34 29074.45 28713.94 29141.59 28614.94
#> 19 29412.07 28855.21 28355.26 28940.68 28921.35
#> 20 29363.64 28426.54 29410.28 28597.90 28606.38
#> 21 29765.27 30040.77 29542.78 30152.22 30183.34
#> 22 32220.66 32057.18 31768.59 32241.53 32183.21
#> 23 34156.03 34057.47 35147.91 34279.69 34235.12
#> 24 35723.96 36689.58 34983.13 36780.37 36654.12
#> 25 36280.06 35648.41 36827.12 35740.49 36103.27
#> 26 33840.14 33668.43 33528.59 33755.07 33530.99
#> 27 34407.08 36135.97 33336.14 36230.09 35938.64
#> 28 32319.76 30719.51 29427.69 30388.45 30291.53
#> 29 29614.98 29097.34 28965.02 28939.13 28766.45
#> 30 27872.45 28863.16 28469.27 28292.87 27561.67
#> 31 29435.29 28800.08 29850.37 28264.59 28077.79
#> 32 29127.54 28455.57 28706.44 27841.23 27636.98
#> 33 29183.36 29158.41 28987.40 29139.26 28900.62
#> 34 31830.06 31279.38 33910.04 31367.97 30857.46
#> 35 32552.70 32838.71 29306.29 33268.83 32517.92
#> 36 34430.62 35206.62 34223.59 34889.89 34586.12
#> 37 35111.78 34116.19 35916.67 33589.01 34260.55
#> 38 32600.07 32144.69 32009.33 32333.50 32024.70
#> 39 32050.98 34506.88 30459.31 34412.39 34143.19
#> 40 28909.21 29035.14 29412.74 29075.38 29027.41
#> 41 28631.67 27933.21 27772.33 27841.73 27936.44
#> 42 27810.99 27664.01 27380.88 27071.34 27063.11
#> 43 27823.04 27418.15 27182.21 27167.99 27457.91
#> 44 27695.70 27062.08 26316.67 26706.98 27127.40
#> 45 27729.13 28251.66 27625.38 27828.42 28191.17
#> 46 30040.28 30452.20 29920.24 30057.55 30284.07
#> 47 31285.80 32283.79 30917.63 31801.27 31998.30
#> 48 33766.69 34779.92 34910.21 34220.15 34461.99
#> 49 36983.56 34822.79 37032.61 34640.33 35186.22
#> 50 32174.73 32478.97 33326.39 32411.67 32321.84
#> 51 35212.61 34396.88 33543.16 34191.48 34098.27
#> 52 28693.48 29109.25 28210.13 28870.24 28995.69
#> 53 27419.96 27808.78 27680.25 27604.85 27791.41
#> 54 27257.66 27042.60 26763.98 26709.01 26698.56
#> 55 26645.51 27175.85 27158.66 26909.30 27167.65
#> 56 26683.68 26725.28 28429.34 26531.20 26783.43
#> 57 27677.94 27991.41 28588.35 27773.89 27935.00
#> 58 29618.71 30345.62 31386.19 30134.59 30162.66
#> 59 31554.72 32120.05 32010.20 31941.42 31863.05
#> 60 34620.52 34622.02 32505.33 34431.48 34257.01
#> 61 35408.60 34167.47 35783.08 33679.08 34171.27
#> 62 30813.78 31214.30 31422.86 30899.38 30993.72
#> 63 32948.11 33573.30 30838.18 33071.76 33032.69
#> 64 27501.02 27831.13 27400.83 27584.47 27886.35
#> 65 28547.37 27270.63 27960.56 26752.54 27308.13
#> 66 27685.63 26759.22 28189.67 26340.52 26516.20
#> 67 27236.53 26857.17 26654.41 26429.01 26937.96
#> 68 27679.89 26823.84 26624.48 26241.10 26716.85
#> 69 27918.96 27702.99 27038.87 27388.42 27752.75
#> 70 28971.84 29941.75 28495.52 29660.45 29843.95
#> 71 30894.37 31637.15 32048.64 31582.71 31650.72
#> 72 32397.33 34187.96 33919.64 34167.32 34091.52
#> 73 33115.75 34016.22 35074.75 33897.86 34124.47
#> 74 32555.78 31310.88 32255.46 31274.89 31335.26
#> 75 31042.54 33283.00 30687.24 32963.94 32871.20
#> 76 30705.99 28457.84 28581.42 28151.62 28412.07
#> 77 28697.64 27759.14 29001.12 27500.41 27822.26
#> 78 26778.47 27158.96 26978.23 26530.16 26623.01
#> 79 26988.92 26860.49 27073.26 26716.07 26928.20
#> 80 27249.51 26564.57 26948.81 26437.79 26605.43
#> 81 27120.25 27427.02 26988.43 27380.79 27435.34
#> 82 30009.34 29802.17 29517.19 29730.54 29642.90
#> 83 31195.88 31354.69 31576.93 31477.45 31247.50
#> 84 32847.38 33889.44 34547.21 34035.28 33683.08
#>
#> $Input_Data$Actual_Test
#> [1] 33043 29574 30718 26992 26681 25357 26047 25762 26293 29030 30305 32410
#> [13] 33543 30547 31497 27266 27081 25786 26269 25865 26495 29046 30206 30558
#> [25] 32910 30755 31315 28613 26586 25174 25548 25506 25711 27117 29988 30748
#> [37] 33248 29049 28653
#>
#> $Input_Data$Forecasts_Test
#> arima ets nnet dampedt dotm
#> [1,] 33395.99 33706.67 33778.28 33846.89 33669.48
#> [2,] 31353.06 30811.06 32761.55 31063.92 30769.82
#> [3,] 32690.41 32715.53 30234.00 32662.11 32104.41
#> [4,] 28822.92 27455.61 28331.94 27445.52 27423.75
#> [5,] 27005.12 26594.99 27446.12 26459.79 26532.87
#> [6,] 25879.31 25478.99 25818.47 25503.73 25395.03
#> [7,] 26128.31 25662.47 26564.64 25634.57 25820.73
#> [8,] 26144.49 25500.46 26877.43 25423.22 25635.85
#> [9,] 26902.03 26541.49 27999.95 26452.49 26554.41
#> [10,] 28515.83 28709.94 28632.77 28761.65 28666.95
#> [11,] 30706.99 30729.40 30382.29 30605.18 30421.59
#> [12,] 30850.00 32834.31 30621.59 32951.20 32606.05
#> [13,] 32665.82 32894.29 32637.60 32915.04 32911.37
#> [14,] 29558.12 30380.45 30544.08 30231.88 30193.74
#> [15,] 30592.16 32128.33 31596.72 31957.74 31738.12
#> [16,] 27364.42 27226.29 28148.55 27102.39 27342.76
#> [17,] 26832.84 26432.22 27447.47 26247.89 26567.58
#> [18,] 25698.38 25476.58 25699.69 25304.02 25472.84
#> [19,] 26272.95 25987.31 26228.02 25633.27 25973.37
#> [20,] 26068.00 25812.51 26082.77 25459.22 25784.94
#> [21,] 26374.39 26600.27 26657.92 26455.79 26659.46
#> [22,] 28708.26 28834.22 29387.03 28829.42 28840.90
#> [23,] 30009.44 30782.86 30667.33 30650.26 30537.55
#> [24,] 31932.58 32854.12 32228.59 32958.64 32680.77
#> [25,] 32061.97 32573.41 32526.32 32805.78 32732.29
#> [26,] 29841.11 30197.23 29858.33 30056.27 29948.50
#> [27,] 31410.68 31662.86 30792.17 31774.25 31509.48
#> [28,] 27873.58 27081.09 28931.22 26959.74 27157.65
#> [29,] 28131.67 26421.91 27213.22 26345.12 26672.68
#> [30,] 26253.58 25669.49 25532.26 25327.80 25477.60
#> [31,] 25708.30 25866.34 26063.55 25602.69 25890.18
#> [32,] 25221.81 25521.81 25442.68 25331.66 25589.83
#> [33,] 25916.13 26382.73 26000.61 26276.04 26409.58
#> [34,] 27976.21 28489.29 28282.95 28595.66 28501.43
#> [35,] 28374.38 29940.13 29089.22 30170.96 29853.79
#> [36,] 29526.02 32015.12 31017.64 32360.05 31894.15
#> [37,] 32539.18 32213.54 32392.26 32468.24 32344.26
#> [38,] 31230.25 29746.24 30849.92 29858.74 29685.34
#> [39,] 30466.33 31211.92 30355.76 31324.26 30923.60
#>
#>
#> $Predict
#> function (object, newpreds)
#> {
#> return(drop(predict(object$lin_model, newpreds, s = "lambda.min")))
#> }
#> <bytecode: 0x7f988e2c2a88>
#> <environment: namespace:ahead>
#>
#> $Intercept
#> [1] 1117.331
#>
#> $Weights
#> [1] 0.02511533 0.00000000 0.26187911 0.00000000 0.66946352
#>
#> $Forecasts_Test
#> [1] 33342.40 31083.61 31348.75 27619.97 26745.93 25529.66 26016.31 25974.86
#> [9] 26902.79 28523.31 30211.18 31739.85 32517.81 30072.16 31407.69 27481.09
#> [17] 26765.19 25546.10 26033.98 25864.65 26608.42 28842.13 30345.94 32237.91
#> [25] 32353.62 29735.50 31064.50 27574.92 26806.83 25519.38 25921.03 25545.15
#> [33] 26257.49 28307.35 29433.84 31333.73 32070.72 29853.89 30534.27
#>
#> $Accuracy_Test
#> ME RMSE MAE MPE MAPE
#> Test set 96.97216 707.1726 531.9225 0.3429199 1.796003
#>
#> $lin_model
#>
#> Call: glmnet::cv.glmnet(x = as.matrix(prediction_matrix), y = as.numeric(observed_vector))
#>
#> Measure: Mean-Squared Error
#>
#> Lambda Index Measure SE Nonzero
#> min 19.3 55 1020679 153117 3
#> 1se 345.3 24 1151856 192213 2
#>
#> attr(,"class")
#> [1] "foreccomb_res" "comb_GLMNET"