Nothing
tests/testthat/_snaps/ALEpDist.md
In ale: Interpretable Machine Learning and Statistical Inference with Accumulated Local Effects (ALE)
ALEpDist works with default inputs (exact) on ALE()
Code
unclass(pd)
Output
<object>
attr(,"S7_class")
<ale::ALEpDist> class
@ parent : <S7_object>
@ constructor: function(model, data, ..., y_col, rand_it, surrogate, parallel, model_packages, random_model_call_string, random_model_call_string_vars, positive, pred_fun, pred_type, output_residuals, seed, silent, .skip_validation) {...}
@ validator : <NULL>
@ properties :
$ rand_stats : <list>
$ residual_distribution: S3<univariateML>
$ residuals : <double> or <NULL>
$ params : <list>
attr(,"rand_stats")
attr(,"rand_stats")$mpg
# A tibble: 10 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 0.000484 -0.00330 0.00201 0 0 0
2 0.00211 -0.00659 0.00835 0.342 -1.56 1.56
3 0.00196 -0.00644 0.00866 0.220 -1.56 1.56
4 0.000908 -0.00363 0.00524 0.0244 -1.56 1.56
5 0.000352 -0.00166 0.00139 0 0 0
6 0.000389 -0.00192 0.00158 0 0 0
7 0.00136 -0.00551 0.00407 0.0977 -1.56 1.56
8 0.000976 -0.00361 0.00715 0.0488 -1.56 1.56
9 0.00280 -0.0136 0.00822 0.439 -1.56 1.56
10 0.000472 -0.00171 0.00149 0 0 0
attr(,"residual_distribution")
Maximum likelihood estimates for the Laplace model
mu sigma
1.303e-11 3.587e-03
attr(,"params")
attr(,"params")$model
attr(,"params")$model$class
[1] "gam" "glm" "lm"
attr(,"params")$model$call
[1] "mgcv::gam(formula = mpg ~ model + s(wt) + am + gear + carb, data = test_cars)"
attr(,"params")$model$print
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nEstimated degrees of freedom:\n8.03 total = 41.03 \n\nGCV score: 0.0001770391 rank: 42/45"
attr(,"params")$model$summary
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nParametric coefficients:\n Estimate Std. Error t value Pr(>|t|) \n(Intercept) 1.432e+01 1.353e-01 105.784 < 2e-16 ***\nmodelCadillac Fleetwood -9.910e+00 1.259e+00 -7.873 5.68e-08 ***\nmodelCamaro Z28 -3.700e+00 7.268e-02 -50.911 < 2e-16 ***\nmodelChrysler Imperial -5.777e+00 1.276e+00 -4.526 0.000152 ***\nmodelDatsun 710 -3.793e+00 1.131e-01 -33.550 < 2e-16 ***\nmodelDodge Challenger -1.266e-01 2.060e-02 -6.147 2.87e-06 ***\nmodelDuster 360 -1.547e+00 2.851e-02 -54.276 < 2e-16 ***\nmodelFerrari Dino -4.088e+00 1.542e-01 -26.506 < 2e-16 ***\nmodelFiat 128 7.211e+00 9.518e-02 75.763 < 2e-16 ***\nmodelFiat X1-9 5.916e+00 1.941e-01 30.488 < 2e-16 ***\nmodelFord Pantera L -1.094e+01 1.737e-01 -63.000 < 2e-16 ***\nmodelHonda Civic 1.474e+01 2.896e-01 50.893 < 2e-16 ***\nmodelHornet 4 Drive 7.569e+00 5.315e-02 142.406 < 2e-16 ***\nmodelHornet Sportabout 3.468e+00 9.616e-03 360.698 < 2e-16 ***\nmodelLincoln Continental -1.023e+01 1.279e+00 -7.998 4.34e-08 ***\nmodelLotus Europa 2.341e+01 3.392e-01 69.015 < 2e-16 ***\nmodelMaserati Bora -1.408e+01 1.903e-01 -74.006 < 2e-16 ***\nmodelMazda RX4 -8.359e+00 1.638e-01 -51.017 < 2e-16 ***\nmodelMazda RX4 Wag -1.030e+01 1.761e-01 -58.494 < 2e-16 ***\nmodelMerc 230 2.481e+00 5.506e-02 45.064 < 2e-16 ***\nmodelMerc 240D 3.804e+00 5.586e-02 68.099 < 2e-16 ***\nmodelMerc 280 -2.984e+00 6.794e-02 -43.926 < 2e-16 ***\nmodelMerc 280C -4.382e+00 6.668e-02 -65.723 < 2e-16 ***\nmodelMerc 450SE -1.661e+00 1.075e-01 -15.448 1.26e-13 ***\nmodelMerc 450SL 7.892e-01 5.311e-02 14.861 2.83e-13 ***\nmodelMerc 450SLC -1.524e+00 6.416e-02 -23.749 < 2e-16 ***\nmodelPontiac Firebird 2.178e+00 7.002e-02 31.102 < 2e-16 ***\nmodelPorsche 914-2 8.306e+00 1.409e-01 58.945 < 2e-16 ***\nmodelToyota Corolla 1.419e+01 2.372e-01 59.809 < 2e-16 ***\nmodelToyota Corona 1.342e+01 2.208e-01 60.795 < 2e-16 ***\nmodelValiant 2.760e+00 1.050e-02 262.897 < 2e-16 ***\nmodelVolvo 142E -9.189e+00 1.720e-01 -53.428 < 2e-16 ***\namTRUE 1.302e+01 1.792e-01 72.629 < 2e-16 ***\ngear.L 1.571e-01 2.703e-02 5.811 6.42e-06 ***\ngear.Q -5.584e+00 4.818e-02 -115.914 < 2e-16 ***\ncarb -3.135e-04 4.119e-03 -0.076 0.939977 \n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nApproximate significance of smooth terms:\n edf Ref.df F p-value \ns(wt) 8.027 8.693 449 <2e-16 ***\n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nRank: 42/45\nR-sq.(adj) = 1 Deviance explained = 100%\nGCV = 0.00017704 Scale est. = 6.3549e-05 n = 64"
attr(,"params")$y_col
[1] "mpg"
attr(,"params")$rand_it
[1] 10
attr(,"params")$parallel
[1] 0
attr(,"params")$model_packages
NULL
attr(,"params")$random_model_call_string
NULL
attr(,"params")$random_model_call_string_vars
character(0)
attr(,"params")$positive
[1] TRUE
attr(,"params")$seed
[1] 0
attr(,"params")$rand_it_ok
[1] 10
attr(,"params")$exactness
[1] "invalid"
Code
unclass(cars_ale)
Output
<object>
attr(,"S7_class")
<ale::ALE> class
@ parent : <S7_object>
@ constructor: function(model, x_cols, data, y_col, ..., exclude_cols, parallel, model_packages, output_stats, output_boot_data, pred_fun, pred_type, p_values, aler_alpha, max_num_bins, boot_it, boot_alpha, boot_centre, seed, y_type, sample_size, silent, .bins) {...}
@ validator : <NULL>
@ properties :
$ effect: <list>
$ params: <list>
attr(,"effect")
attr(,"effect")$mpg
attr(,"effect")$mpg$ale
attr(,"effect")$mpg$ale$d1
attr(,"effect")$mpg$ale$d1$vs
# A tibble: 2 x 7
vs.bin .n .y .y_lo .y_mean .y_median .y_hi
<ord> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 FALSE 36 0 0 0 0 0
2 TRUE 28 0 0 0 0 0
attr(,"effect")$mpg$ale$d1$continent
# A tibble: 3 x 7
continent.bin .n .y .y_lo .y_mean .y_median .y_hi
<ord> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 North America 24 0 0 0 0 0
2 Europe 28 0 0 0 0 0
3 Asia 12 0 0 0 0 0
attr(,"effect")$mpg$ale$d1$am
# A tibble: 2 x 7
am.bin .n .y .y_lo .y_mean .y_median .y_hi
<ord> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 FALSE 38 -1.61 -5.01 -1.61 -1.01 0.775
2 TRUE 26 1.60 -1.75 1.60 0.219 7.28
attr(,"effect")$mpg$ale$d1$model
# A tibble: 32 x 7
model.bin .n .y .y_lo .y_mean .y_median .y_hi
<ord> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 Camaro Z28 2 -1.22 -3.83 -1.22 -1.74 2.27
2 Cadillac Fleetwood 2 -0.330 -9.42 -0.330 0.233 7.80
3 Lincoln Continental 2 7.05 -9.18 7.05 7.26 22.9
4 Chrysler Imperial 2 15.5 -4.52 15.5 14.3 37.4
5 Duster 360 2 21.7 -0.182 21.7 21.6 43.9
6 Hornet Sportabout 2 31.8 5.16 31.8 37.6 48.4
7 Pontiac Firebird 2 25.1 3.36 25.1 29.3 39.7
8 Dodge Challenger 2 20.3 0.910 20.3 21.6 37.4
9 AMC Javelin 2 20.4 1.16 20.4 22.0 36.9
10 Merc 450SL 2 20.8 1.13 20.8 15.1 50.2
# i 22 more rows
attr(,"effect")$mpg$ale$d1$gear
# A tibble: 3 x 7
gear.bin .n .y .y_lo .y_mean .y_median .y_hi
<ord> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 three 30 -1.06 -2.54 -1.06 -0.758 -0.0935
2 four 24 2.66 1.49 2.66 2.42 4.22
3 five 10 -1.60 -2.46 -1.60 -1.85 -0.331
attr(,"effect")$mpg$ale$d1$carb
# A tibble: 5 x 7
carb.ceil .n .y .y_lo .y_mean .y_median .y_hi
<dbl> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 1 14 0.000490 0.000490 0.000490 0.000490 0.000490
2 2 19 0.000176 0.000176 0.000176 0.000176 0.000176
3 3 9 -0.000137 -0.000137 -0.000137 -0.000137 -0.000137
4 4 16 -0.000451 -0.000451 -0.000451 -0.000451 -0.000451
5 8 6 -0.00170 -0.00170 -0.00170 -0.00170 -0.00170
attr(,"effect")$mpg$ale$d1$wt
# A tibble: 10 x 7
wt.ceil .n .y .y_lo .y_mean .y_median .y_hi
<dbl> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 1.50 1 -18.1 -18.1 -18.1 -18.1 -18.1
2 1.94 7 -10.2 -10.2 -10.2 -10.2 -10.2
3 2.46 7 -3.51 -3.51 -3.51 -3.51 -3.51
4 2.79 7 -0.882 -0.882 -0.882 -0.882 -0.882
5 3.19 7 2.07 2.07 2.07 2.07 2.07
6 3.44 7 3.65 3.65 3.65 3.65 3.65
7 3.52 7 4.05 4.05 4.05 4.05 4.05
8 3.73 7 4.92 4.92 4.92 4.92 4.92
9 4.05 7 6.38 6.38 6.38 6.38 6.38
10 5.45 7 9.11 9.11 9.11 9.11 9.11
attr(,"effect")$mpg$stats
attr(,"effect")$mpg$stats$d1
# A tibble: 42 x 8
statistic estimate p.value term conf.low mean median conf.high
<chr> <dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 aled 0 1 vs 0 0 0 0
2 aler_min 0 1 vs 0 0 0 0
3 aler_max 0 1 vs 0 0 0 0
4 naled 0 0.6 vs 0 0 0 0
5 naler_min 0 1 vs 0 0 0 0
6 naler_max 0 0.6 vs 0 0 0 0
7 aled 0 1 continent 0 0 0 0
8 aler_min 0 1 continent 0 0 0 0
9 aler_max 0 1 continent 0 0 0 0
10 naled 0 0.6 continent 0 0 0 0
# i 32 more rows
attr(,"effect")$mpg$boot_data
NULL
attr(,"params")
attr(,"params")$max_d
[1] 1
attr(,"params")$ordered_x_cols
attr(,"params")$ordered_x_cols$d1
[1] "vs" "continent" "am" "model" "gear" "carb"
[7] "wt"
attr(,"params")$ordered_x_cols$d2
character(0)
attr(,"params")$requested_x_cols
attr(,"params")$requested_x_cols$d1
[1] "vs" "continent" "am" "model" "gear" "carb"
[7] "wt"
attr(,"params")$requested_x_cols$d2
character(0)
attr(,"params")$y_cats
[1] "mpg"
attr(,"params")$y_summary
mpg
min 10.39108
1% 10.39108
2.5% 10.40000
5% 10.88271
10% 14.33418
20% 15.16500
25% 15.43921
30% 15.79628
40% 17.83840
aler_lo_lo 19.18669
aler_lo 19.18797
50% 19.20000
mean 20.09462
aler_hi 19.20859
aler_hi_hi 19.20864
60% 21.00000
70% 21.51193
75% 22.80000
80% 24.48680
90% 30.31124
95% 32.14486
97.5% 33.08402
99% 33.84876
max 33.84876
attr(,"params")$model
attr(,"params")$model$class
[1] "gam" "glm" "lm"
attr(,"params")$model$call
[1] "mgcv::gam(formula = mpg ~ model + s(wt) + am + gear + carb, data = test_cars)"
attr(,"params")$model$print
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nEstimated degrees of freedom:\n8.03 total = 41.03 \n\nGCV score: 0.0001770391 rank: 42/45"
attr(,"params")$model$summary
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nParametric coefficients:\n Estimate Std. Error t value Pr(>|t|) \n(Intercept) 1.432e+01 1.353e-01 105.784 < 2e-16 ***\nmodelCadillac Fleetwood -9.910e+00 1.259e+00 -7.873 5.68e-08 ***\nmodelCamaro Z28 -3.700e+00 7.268e-02 -50.911 < 2e-16 ***\nmodelChrysler Imperial -5.777e+00 1.276e+00 -4.526 0.000152 ***\nmodelDatsun 710 -3.793e+00 1.131e-01 -33.550 < 2e-16 ***\nmodelDodge Challenger -1.266e-01 2.060e-02 -6.147 2.87e-06 ***\nmodelDuster 360 -1.547e+00 2.851e-02 -54.276 < 2e-16 ***\nmodelFerrari Dino -4.088e+00 1.542e-01 -26.506 < 2e-16 ***\nmodelFiat 128 7.211e+00 9.518e-02 75.763 < 2e-16 ***\nmodelFiat X1-9 5.916e+00 1.941e-01 30.488 < 2e-16 ***\nmodelFord Pantera L -1.094e+01 1.737e-01 -63.000 < 2e-16 ***\nmodelHonda Civic 1.474e+01 2.896e-01 50.893 < 2e-16 ***\nmodelHornet 4 Drive 7.569e+00 5.315e-02 142.406 < 2e-16 ***\nmodelHornet Sportabout 3.468e+00 9.616e-03 360.698 < 2e-16 ***\nmodelLincoln Continental -1.023e+01 1.279e+00 -7.998 4.34e-08 ***\nmodelLotus Europa 2.341e+01 3.392e-01 69.015 < 2e-16 ***\nmodelMaserati Bora -1.408e+01 1.903e-01 -74.006 < 2e-16 ***\nmodelMazda RX4 -8.359e+00 1.638e-01 -51.017 < 2e-16 ***\nmodelMazda RX4 Wag -1.030e+01 1.761e-01 -58.494 < 2e-16 ***\nmodelMerc 230 2.481e+00 5.506e-02 45.064 < 2e-16 ***\nmodelMerc 240D 3.804e+00 5.586e-02 68.099 < 2e-16 ***\nmodelMerc 280 -2.984e+00 6.794e-02 -43.926 < 2e-16 ***\nmodelMerc 280C -4.382e+00 6.668e-02 -65.723 < 2e-16 ***\nmodelMerc 450SE -1.661e+00 1.075e-01 -15.448 1.26e-13 ***\nmodelMerc 450SL 7.892e-01 5.311e-02 14.861 2.83e-13 ***\nmodelMerc 450SLC -1.524e+00 6.416e-02 -23.749 < 2e-16 ***\nmodelPontiac Firebird 2.178e+00 7.002e-02 31.102 < 2e-16 ***\nmodelPorsche 914-2 8.306e+00 1.409e-01 58.945 < 2e-16 ***\nmodelToyota Corolla 1.419e+01 2.372e-01 59.809 < 2e-16 ***\nmodelToyota Corona 1.342e+01 2.208e-01 60.795 < 2e-16 ***\nmodelValiant 2.760e+00 1.050e-02 262.897 < 2e-16 ***\nmodelVolvo 142E -9.189e+00 1.720e-01 -53.428 < 2e-16 ***\namTRUE 1.302e+01 1.792e-01 72.629 < 2e-16 ***\ngear.L 1.571e-01 2.703e-02 5.811 6.42e-06 ***\ngear.Q -5.584e+00 4.818e-02 -115.914 < 2e-16 ***\ncarb -3.135e-04 4.119e-03 -0.076 0.939977 \n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nApproximate significance of smooth terms:\n edf Ref.df F p-value \ns(wt) 8.027 8.693 449 <2e-16 ***\n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nRank: 42/45\nR-sq.(adj) = 1 Deviance explained = 100%\nGCV = 0.00017704 Scale est. = 6.3549e-05 n = 64"
attr(,"params")$data
attr(,"params")$data$data_sample
# A tibble: 64 x 8
mpg vs continent am model gear carb wt
<dbl> <lgl> <fct> <lgl> <chr> <ord> <int> <dbl>
1 21 FALSE Asia TRUE Mazda RX4 four 4 2.62
2 21 FALSE Asia TRUE Mazda RX4 Wag four 4 2.88
3 22.8 TRUE Asia TRUE Datsun 710 four 1 2.32
4 21.4 TRUE North America FALSE Hornet 4 Drive three 1 3.22
5 18.7 FALSE North America FALSE Hornet Sportabout three 2 3.44
6 18.1 TRUE North America FALSE Valiant three 1 3.46
7 14.3 FALSE North America FALSE Duster 360 three 4 3.57
8 24.4 TRUE Europe FALSE Merc 240D four 2 3.19
9 22.8 TRUE Europe FALSE Merc 230 four 2 3.15
10 19.2 TRUE Europe FALSE Merc 280 four 4 3.44
# i 54 more rows
attr(,"params")$data$y_vals_sample
mpg
[1,] 21.00000
[2,] 21.00000
[3,] 22.80000
[4,] 21.40000
[5,] 18.70000
[6,] 18.10000
[7,] 14.30000
[8,] 24.40000
[9,] 22.80000
[10,] 19.20000
[11,] 17.80000
[12,] 16.40000
[13,] 17.30000
[14,] 15.20000
[15,] 10.40000
[16,] 10.40000
[17,] 14.70000
[18,] 32.40000
[19,] 30.40000
[20,] 33.90000
[21,] 21.50000
[22,] 15.50000
[23,] 15.20000
[24,] 13.30000
[25,] 19.20000
[26,] 27.30000
[27,] 26.00000
[28,] 30.40000
[29,] 15.80000
[30,] 19.70000
[31,] 15.00000
[32,] 21.40000
[33,] 21.16661
[34,] 20.90151
[35,] 22.74169
[36,] 21.43118
[37,] 18.85267
[38,] 17.99201
[39,] 14.41394
[40,] 24.61700
[41,] 22.87332
[42,] 19.24958
[43,] 17.64400
[44,] 16.30356
[45,] 17.18809
[46,] 15.25685
[47,] 10.37589
[48,] 10.45613
[49,] 14.69932
[50,] 32.54102
[51,] 30.69908
[52,] 33.81866
[53,] 21.61930
[54,] 15.63476
[55,] 15.11249
[56,] 13.34035
[57,] 19.05621
[58,] 27.17290
[59,] 25.94078
[60,] 30.10414
[61,] 15.76283
[62,] 19.84566
[63,] 14.95210
[64,] 21.39233
attr(,"params")$data$nrow
[1] 64
attr(,"params")$y_col
[1] "mpg"
attr(,"params")$parallel
[1] 0
attr(,"params")$model_packages
NULL
attr(,"params")$output_stats
[1] TRUE
attr(,"params")$output_boot_data
[1] FALSE
attr(,"params")$pred_fun
[1] "function(object, newdata, type = pred_type) {\n stats::predict(object = object, newdata = newdata, type = type)\n }"
attr(,"params")$pred_type
[1] "response"
attr(,"params")$p_values
<ale::ALEpDist>
@ rand_stats :List of 1
.. $ mpg: tibble [10 x 6] (S3: tbl_df/tbl/data.frame)
.. ..$ aled : num [1:10] 0.000484 0.002108 0.001961 0.000908 0.000352 ...
.. ..$ aler_min : num [1:10] -0.0033 -0.00659 -0.00644 -0.00363 -0.00166 ...
.. ..$ aler_max : num [1:10] 0.00201 0.00835 0.00866 0.00524 0.00139 ...
.. ..$ naled : num [1:10] 0 0.3418 0.2197 0.0244 0 ...
.. ..$ naler_min: num [1:10] 0 -1.56 -1.56 -1.56 0 ...
.. ..$ naler_max: num [1:10] 0 1.56 1.56 1.56 0 ...
@ residual_distribution: 'univariateML' Named num [1:2] 1.30e-11 3.59e-03
.. - attr(*, "logLik")= num 252
.. - attr(*, "call")= language f(x = x, na.rm = na.rm)
.. - attr(*, "n")= int 64
.. - attr(*, "model")= chr "Laplace"
.. - attr(*, "density")= chr "extraDistr::dlaplace"
.. - attr(*, "support")= num [1:2] -Inf Inf
.. - attr(*, "names")= chr [1:2] "mu" "sigma"
.. - attr(*, "default")= num [1:2] 0 1
.. - attr(*, "continuous")= logi TRUE
@ residuals : NULL
@ params :List of 11
.. $ model :List of 4
.. ..$ class : chr [1:3] "gam" "glm" "lm"
.. ..$ call : chr "mgcv::gam(formula = mpg ~ model + s(wt) + am + gear + carb, data = test_cars)"
.. ..$ print : chr "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nEstimated "| __truncated__
.. ..$ summary: chr "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nParametric"| __truncated__
.. $ y_col : chr "mpg"
.. $ rand_it : num 10
.. $ parallel : num 0
.. $ model_packages : NULL
.. $ random_model_call_string : NULL
.. $ random_model_call_string_vars: chr(0)
.. $ positive : logi TRUE
.. $ seed : num 0
.. $ rand_it_ok : int 10
.. $ exactness : chr "invalid"
attr(,"params")$aler_alpha
[1] 0.01 0.05
attr(,"params")$max_num_bins
[1] 10
attr(,"params")$boot_it
[1] 3
attr(,"params")$boot_alpha
[1] 0.05
attr(,"params")$boot_centre
[1] "mean"
attr(,"params")$seed
[1] 0
attr(,"params")$y_type
[1] "numeric"
attr(,"params")$sample_size
[1] 500
Code
unclass(set_names(map(stats_names, function(.stat) {
value_to_p(pd@rand_stats$mpg, .stat, test_vals)
}), stats_names))
Output
$aled
[1] 1 1 1 1 1 0 0 0 0 0 0
$aler_min
[1] 0 0 0 0 1 1 1 1 1 1 1
$aler_max
[1] 1 1 1 1 1 0 0 0 0 0 0
$naled
[1] 1.0 1.0 1.0 1.0 0.6 0.4 0.3 0.0 0.0 0.0 0.0
$naler_min
[1] 0.0 0.0 0.6 0.6 1.0 1.0 1.0 1.0 1.0 1.0 1.0
$naler_max
[1] 1.0 1.0 1.0 1.0 0.6 0.6 0.6 0.6 0.6 0.0 0.0
Surrogate ALEpDist works
Code
unclass(pd)
Output
<object>
attr(,"S7_class")
<ale::ALEpDist> class
@ parent : <S7_object>
@ constructor: function(model, data, ..., y_col, rand_it, surrogate, parallel, model_packages, random_model_call_string, random_model_call_string_vars, positive, pred_fun, pred_type, output_residuals, seed, silent, .skip_validation) {...}
@ validator : <NULL>
@ properties :
$ rand_stats : <list>
$ residual_distribution: S3<univariateML>
$ residuals : <double> or <NULL>
$ params : <list>
attr(,"rand_stats")
attr(,"rand_stats")$mpg
# A tibble: 100 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 0.00199 -0.0112 0.00727 0.366 -1.56 1.56
2 0.00394 -0.0118 0.0148 0.708 -1.56 1.56
3 0.00425 -0.0170 0.0126 0.708 -1.56 1.56
4 0.00647 -0.0215 0.0304 0.708 -1.56 1.56
5 0.00365 -0.0149 0.0124 0.537 -1.56 1.56
6 0.00281 -0.0121 0.0100 0.366 -1.56 1.56
7 0.00599 -0.0172 0.0227 0.879 -1.56 1.56
8 0.00574 -0.0333 0.0175 0.708 -1.56 1.56
9 0.00733 -0.0321 0.0203 0.879 -1.56 1.56
10 0.00729 -0.0228 0.0257 1.05 -1.56 1.56
# i 90 more rows
attr(,"residual_distribution")
Maximum likelihood estimates for the Laplace model
mu sigma
1.303e-11 3.587e-03
attr(,"residuals")
[1] -9.470698e-04 -1.130145e-03 -3.078035e-03 7.415332e-04 -4.678952e-03
[6] 7.516518e-04 2.728091e-03 -8.853029e-03 -3.016706e-04 -1.794893e-03
[11] -3.673897e-03 -2.816578e-03 5.414042e-03 2.979146e-03 2.219206e-03
[16] 8.324011e-04 -4.976941e-05 -1.115543e-02 1.437735e-03 2.200997e-03
[21] 2.625747e-03 5.178720e-04 -9.802341e-03 7.118944e-03 5.255702e-03
[26] -9.746617e-03 -2.976337e-03 6.542735e-03 -8.071930e-03 4.016990e-03
[31] -2.747836e-04 -6.032700e-05 9.470698e-04 1.130145e-03 3.078035e-03
[36] -7.415332e-04 4.678952e-03 -7.516518e-04 -2.728091e-03 8.853029e-03
[41] 3.016706e-04 1.794893e-03 3.673897e-03 2.816578e-03 -5.414042e-03
[46] -2.979146e-03 -2.219206e-03 -8.324011e-04 4.976944e-05 1.115543e-02
[51] -1.437735e-03 -2.200997e-03 -2.625747e-03 -5.178720e-04 9.802341e-03
[56] -7.118944e-03 -5.255702e-03 9.746617e-03 2.976337e-03 -6.542735e-03
[61] 8.071930e-03 -4.016990e-03 2.747836e-04 6.032702e-05
attr(,"params")
attr(,"params")$model
attr(,"params")$model$class
[1] "gam" "glm" "lm"
attr(,"params")$model$call
[1] "mgcv::gam(formula = mpg ~ model + s(wt) + am + gear + carb, data = test_cars)"
attr(,"params")$model$print
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nEstimated degrees of freedom:\n8.03 total = 41.03 \n\nGCV score: 0.0001770391 rank: 42/45"
attr(,"params")$model$summary
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nParametric coefficients:\n Estimate Std. Error t value Pr(>|t|) \n(Intercept) 1.432e+01 1.353e-01 105.784 < 2e-16 ***\nmodelCadillac Fleetwood -9.910e+00 1.259e+00 -7.873 5.68e-08 ***\nmodelCamaro Z28 -3.700e+00 7.268e-02 -50.911 < 2e-16 ***\nmodelChrysler Imperial -5.777e+00 1.276e+00 -4.526 0.000152 ***\nmodelDatsun 710 -3.793e+00 1.131e-01 -33.550 < 2e-16 ***\nmodelDodge Challenger -1.266e-01 2.060e-02 -6.147 2.87e-06 ***\nmodelDuster 360 -1.547e+00 2.851e-02 -54.276 < 2e-16 ***\nmodelFerrari Dino -4.088e+00 1.542e-01 -26.506 < 2e-16 ***\nmodelFiat 128 7.211e+00 9.518e-02 75.763 < 2e-16 ***\nmodelFiat X1-9 5.916e+00 1.941e-01 30.488 < 2e-16 ***\nmodelFord Pantera L -1.094e+01 1.737e-01 -63.000 < 2e-16 ***\nmodelHonda Civic 1.474e+01 2.896e-01 50.893 < 2e-16 ***\nmodelHornet 4 Drive 7.569e+00 5.315e-02 142.406 < 2e-16 ***\nmodelHornet Sportabout 3.468e+00 9.616e-03 360.698 < 2e-16 ***\nmodelLincoln Continental -1.023e+01 1.279e+00 -7.998 4.34e-08 ***\nmodelLotus Europa 2.341e+01 3.392e-01 69.015 < 2e-16 ***\nmodelMaserati Bora -1.408e+01 1.903e-01 -74.006 < 2e-16 ***\nmodelMazda RX4 -8.359e+00 1.638e-01 -51.017 < 2e-16 ***\nmodelMazda RX4 Wag -1.030e+01 1.761e-01 -58.494 < 2e-16 ***\nmodelMerc 230 2.481e+00 5.506e-02 45.064 < 2e-16 ***\nmodelMerc 240D 3.804e+00 5.586e-02 68.099 < 2e-16 ***\nmodelMerc 280 -2.984e+00 6.794e-02 -43.926 < 2e-16 ***\nmodelMerc 280C -4.382e+00 6.668e-02 -65.723 < 2e-16 ***\nmodelMerc 450SE -1.661e+00 1.075e-01 -15.448 1.26e-13 ***\nmodelMerc 450SL 7.892e-01 5.311e-02 14.861 2.83e-13 ***\nmodelMerc 450SLC -1.524e+00 6.416e-02 -23.749 < 2e-16 ***\nmodelPontiac Firebird 2.178e+00 7.002e-02 31.102 < 2e-16 ***\nmodelPorsche 914-2 8.306e+00 1.409e-01 58.945 < 2e-16 ***\nmodelToyota Corolla 1.419e+01 2.372e-01 59.809 < 2e-16 ***\nmodelToyota Corona 1.342e+01 2.208e-01 60.795 < 2e-16 ***\nmodelValiant 2.760e+00 1.050e-02 262.897 < 2e-16 ***\nmodelVolvo 142E -9.189e+00 1.720e-01 -53.428 < 2e-16 ***\namTRUE 1.302e+01 1.792e-01 72.629 < 2e-16 ***\ngear.L 1.571e-01 2.703e-02 5.811 6.42e-06 ***\ngear.Q -5.584e+00 4.818e-02 -115.914 < 2e-16 ***\ncarb -3.135e-04 4.119e-03 -0.076 0.939977 \n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nApproximate significance of smooth terms:\n edf Ref.df F p-value \ns(wt) 8.027 8.693 449 <2e-16 ***\n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nRank: 42/45\nR-sq.(adj) = 1 Deviance explained = 100%\nGCV = 0.00017704 Scale est. = 6.3549e-05 n = 64"
attr(,"params")$y_col
[1] "mpg"
attr(,"params")$rand_it
[1] 3
attr(,"params")$parallel
[1] 0
attr(,"params")$model_packages
NULL
attr(,"params")$random_model_call_string
NULL
attr(,"params")$random_model_call_string_vars
character(0)
attr(,"params")$positive
[1] TRUE
attr(,"params")$seed
[1] 0
attr(,"params")$rand_it_ok
[1] 100
attr(,"params")$exactness
[1] "surrogate"
Code
unclass(set_names(map(stats_names, function(.stat) {
p_to_random_value(pd@rand_stats$mpg, .stat, test_p)
}), stats_names))
Output
$aled
0 0.001 0.01 0.01 0.05 0.1
2.145276e-02 2.092763e-02 1.620142e-02 1.620142e-02 1.331799e-02 1.160576e-02
0.5 1
5.855495e-03 2.343212e-05
$aler_min
0 0.001 0.01 0.01 0.05
-8.018319e-02 -7.889883e-02 -6.733957e-02 -6.733957e-02 -4.856189e-02
0.1 0.5 1
-4.264222e-02 -1.985118e-02 -8.389436e-05
$aler_max
0 0.001 0.01 0.01 0.05 0.1
6.705638e-02 6.689973e-02 6.548988e-02 6.548988e-02 5.404877e-02 5.006009e-02
0.5 1
2.010266e-02 5.471472e-05
$naled
0 0.001 0.01 0.01 0.05 0.1 0.5 1
1.3916016 1.3916016 1.3916016 1.3916016 1.2207031 1.2207031 0.7080078 0.0000000
$naler_min
0 0.001 0.01 0.01 0.05 0.1 0.5 1
-1.5625 -1.5625 -1.5625 -1.5625 -1.5625 -1.5625 -1.5625 0.0000
$naler_max
0 0.001 0.01 0.01 0.05 0.1 0.5 1
3.1250 3.1250 3.1250 3.1250 3.1250 3.1250 1.5625 0.0000
ALEpDist works with custom random_model_call_string
Code
unclass(pd)
Output
<object>
attr(,"S7_class")
<ale::ALEpDist> class
@ parent : <S7_object>
@ constructor: function(model, data, ..., y_col, rand_it, surrogate, parallel, model_packages, random_model_call_string, random_model_call_string_vars, positive, pred_fun, pred_type, output_residuals, seed, silent, .skip_validation) {...}
@ validator : <NULL>
@ properties :
$ rand_stats : <list>
$ residual_distribution: S3<univariateML>
$ residuals : <double> or <NULL>
$ params : <list>
attr(,"rand_stats")
attr(,"rand_stats")$mpg
# A tibble: 3 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 0.000484 -0.00330 0.00201 0 0 0
2 0.00211 -0.00659 0.00835 0.342 -1.56 1.56
3 0.00196 -0.00644 0.00866 0.220 -1.56 1.56
attr(,"residual_distribution")
Maximum likelihood estimates for the Laplace model
mu sigma
1.303e-11 3.587e-03
attr(,"residuals")
[1] -9.470698e-04 -1.130145e-03 -3.078035e-03 7.415332e-04 -4.678952e-03
[6] 7.516518e-04 2.728091e-03 -8.853029e-03 -3.016706e-04 -1.794893e-03
[11] -3.673897e-03 -2.816578e-03 5.414042e-03 2.979146e-03 2.219206e-03
[16] 8.324011e-04 -4.976941e-05 -1.115543e-02 1.437735e-03 2.200997e-03
[21] 2.625747e-03 5.178720e-04 -9.802341e-03 7.118944e-03 5.255702e-03
[26] -9.746617e-03 -2.976337e-03 6.542735e-03 -8.071930e-03 4.016990e-03
[31] -2.747836e-04 -6.032700e-05 9.470698e-04 1.130145e-03 3.078035e-03
[36] -7.415332e-04 4.678952e-03 -7.516518e-04 -2.728091e-03 8.853029e-03
[41] 3.016706e-04 1.794893e-03 3.673897e-03 2.816578e-03 -5.414042e-03
[46] -2.979146e-03 -2.219206e-03 -8.324011e-04 4.976944e-05 1.115543e-02
[51] -1.437735e-03 -2.200997e-03 -2.625747e-03 -5.178720e-04 9.802341e-03
[56] -7.118944e-03 -5.255702e-03 9.746617e-03 2.976337e-03 -6.542735e-03
[61] 8.071930e-03 -4.016990e-03 2.747836e-04 6.032702e-05
attr(,"params")
attr(,"params")$model
attr(,"params")$model$class
[1] "gam" "glm" "lm"
attr(,"params")$model$call
[1] "mgcv::gam(formula = mpg ~ model + s(wt) + am + gear + carb, data = test_cars)"
attr(,"params")$model$print
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nEstimated degrees of freedom:\n8.03 total = 41.03 \n\nGCV score: 0.0001770391 rank: 42/45"
attr(,"params")$model$summary
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nParametric coefficients:\n Estimate Std. Error t value Pr(>|t|) \n(Intercept) 1.432e+01 1.353e-01 105.784 < 2e-16 ***\nmodelCadillac Fleetwood -9.910e+00 1.259e+00 -7.873 5.68e-08 ***\nmodelCamaro Z28 -3.700e+00 7.268e-02 -50.911 < 2e-16 ***\nmodelChrysler Imperial -5.777e+00 1.276e+00 -4.526 0.000152 ***\nmodelDatsun 710 -3.793e+00 1.131e-01 -33.550 < 2e-16 ***\nmodelDodge Challenger -1.266e-01 2.060e-02 -6.147 2.87e-06 ***\nmodelDuster 360 -1.547e+00 2.851e-02 -54.276 < 2e-16 ***\nmodelFerrari Dino -4.088e+00 1.542e-01 -26.506 < 2e-16 ***\nmodelFiat 128 7.211e+00 9.518e-02 75.763 < 2e-16 ***\nmodelFiat X1-9 5.916e+00 1.941e-01 30.488 < 2e-16 ***\nmodelFord Pantera L -1.094e+01 1.737e-01 -63.000 < 2e-16 ***\nmodelHonda Civic 1.474e+01 2.896e-01 50.893 < 2e-16 ***\nmodelHornet 4 Drive 7.569e+00 5.315e-02 142.406 < 2e-16 ***\nmodelHornet Sportabout 3.468e+00 9.616e-03 360.698 < 2e-16 ***\nmodelLincoln Continental -1.023e+01 1.279e+00 -7.998 4.34e-08 ***\nmodelLotus Europa 2.341e+01 3.392e-01 69.015 < 2e-16 ***\nmodelMaserati Bora -1.408e+01 1.903e-01 -74.006 < 2e-16 ***\nmodelMazda RX4 -8.359e+00 1.638e-01 -51.017 < 2e-16 ***\nmodelMazda RX4 Wag -1.030e+01 1.761e-01 -58.494 < 2e-16 ***\nmodelMerc 230 2.481e+00 5.506e-02 45.064 < 2e-16 ***\nmodelMerc 240D 3.804e+00 5.586e-02 68.099 < 2e-16 ***\nmodelMerc 280 -2.984e+00 6.794e-02 -43.926 < 2e-16 ***\nmodelMerc 280C -4.382e+00 6.668e-02 -65.723 < 2e-16 ***\nmodelMerc 450SE -1.661e+00 1.075e-01 -15.448 1.26e-13 ***\nmodelMerc 450SL 7.892e-01 5.311e-02 14.861 2.83e-13 ***\nmodelMerc 450SLC -1.524e+00 6.416e-02 -23.749 < 2e-16 ***\nmodelPontiac Firebird 2.178e+00 7.002e-02 31.102 < 2e-16 ***\nmodelPorsche 914-2 8.306e+00 1.409e-01 58.945 < 2e-16 ***\nmodelToyota Corolla 1.419e+01 2.372e-01 59.809 < 2e-16 ***\nmodelToyota Corona 1.342e+01 2.208e-01 60.795 < 2e-16 ***\nmodelValiant 2.760e+00 1.050e-02 262.897 < 2e-16 ***\nmodelVolvo 142E -9.189e+00 1.720e-01 -53.428 < 2e-16 ***\namTRUE 1.302e+01 1.792e-01 72.629 < 2e-16 ***\ngear.L 1.571e-01 2.703e-02 5.811 6.42e-06 ***\ngear.Q -5.584e+00 4.818e-02 -115.914 < 2e-16 ***\ncarb -3.135e-04 4.119e-03 -0.076 0.939977 \n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nApproximate significance of smooth terms:\n edf Ref.df F p-value \ns(wt) 8.027 8.693 449 <2e-16 ***\n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nRank: 42/45\nR-sq.(adj) = 1 Deviance explained = 100%\nGCV = 0.00017704 Scale est. = 6.3549e-05 n = 64"
attr(,"params")$y_col
[1] "mpg"
attr(,"params")$rand_it
[1] 3
attr(,"params")$parallel
[1] 0
attr(,"params")$model_packages
NULL
attr(,"params")$random_model_call_string
[1] "mgcv::gam(\n mpg ~ model + s(wt) + am + gear + carb + random_variable,\n data = it.rand_data\n )"
attr(,"params")$random_model_call_string_vars
[1] "rmcsv"
attr(,"params")$positive
[1] TRUE
attr(,"params")$seed
[1] 0
attr(,"params")$rand_it_ok
[1] 3
attr(,"params")$exactness
[1] "invalid"
ALEpDist works with binary outcome
Code
unclass(pd)
Output
<object>
attr(,"S7_class")
<ale::ALEpDist> class
@ parent : <S7_object>
@ constructor: function(model, data, ..., y_col, rand_it, surrogate, parallel, model_packages, random_model_call_string, random_model_call_string_vars, positive, pred_fun, pred_type, output_residuals, seed, silent, .skip_validation) {...}
@ validator : <NULL>
@ properties :
$ rand_stats : <list>
$ residual_distribution: S3<univariateML>
$ residuals : <double> or <NULL>
$ params : <list>
attr(,"rand_stats")
attr(,"rand_stats")$vs
# A tibble: 10 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 5.16e-25 -1.32e-24 9.89e-25 3.64 -7.81 6.25
2 2.70e-25 -6.53e-25 6.73e-25 2.56 -1.56 4.69
3 1.43e-25 -3.61e-25 3.58e-25 2.22 -1.56 4.69
4 2.86e-25 -5.04e-25 4.75e-25 2.81 -1.56 4.69
5 8.00e-23 -1.39e-22 1.57e-22 23.8 -50 6.25
6 8.44e-26 -2.01e-25 1.96e-25 2.17 -1.56 3.12
7 1.12e-23 -2.59e-23 2.06e-23 22.8 -50 6.25
8 1.28e-21 -2.04e-21 2.84e-21 26.1 -50 6.25
9 0 0 0 0 0 0
10 6.51e-24 -1.16e-23 1.15e-23 19.7 -50 6.25
attr(,"residual_distribution")
Maximum likelihood estimates for the Uniform model
min max
-3.926e-13 3.926e-13
attr(,"params")
attr(,"params")$model
attr(,"params")$model$class
[1] "gam" "glm" "lm"
attr(,"params")$model$call
[1] "mgcv::gam(formula = vs ~ model + s(wt) + am + gear + carb, family = stats::binomial(), \n data = test_cars)"
attr(,"params")$model$print
[1] "\nFamily: binomial \nLink function: logit \n\nFormula:\nvs ~ model + s(wt) + am + gear + carb\n\nEstimated degrees of freedom:\n1 total = 34 \n\nUBRE score: 0.0625 rank: 42/45"
attr(,"params")$model$summary
[1] "\nFamily: binomial \nLink function: logit \n\nFormula:\nvs ~ model + s(wt) + am + gear + carb\n\nParametric coefficients:\n Estimate Std. Error z value Pr(>|z|)\n(Intercept) 9.522e+00 1.068e+06 0 1\nmodelCadillac Fleetwood 1.310e-09 2.420e+07 0 1\nmodelCamaro Z28 2.970e-10 5.648e+06 0 1\nmodelChrysler Imperial 1.380e-09 2.550e+07 0 1\nmodelDatsun 710 7.775e-08 9.376e+06 0 1\nmodelDodge Challenger 7.967e-11 1.760e+06 0 1\nmodelDuster 360 1.078e-10 2.390e+06 0 1\nmodelFerrari Dino -5.713e+01 1.689e+07 0 1\nmodelFiat 128 4.992e-08 7.891e+06 0 1\nmodelFiat X1-9 8.312e-08 4.293e+06 0 1\nmodelFord Pantera L -5.713e+01 2.207e+07 0 1\nmodelHonda Civic 0.000e+00 0.000e+00 NaN NaN\nmodelHornet 4 Drive 5.713e+01 2.914e+06 0 1\nmodelHornet Sportabout 1.906e-11 1.054e+06 0 1\nmodelLincoln Continental 1.451e-09 2.680e+07 0 1\nmodelLotus Europa 0.000e+00 0.000e+00 NaN NaN\nmodelMaserati Bora -5.713e+01 2.726e+07 0 1\nmodelMazda RX4 -5.713e+01 1.339e+07 0 1\nmodelMazda RX4 Wag -5.713e+01 1.657e+07 0 1\nmodelMerc 230 -1.764e-05 1.169e+06 0 1\nmodelMerc 240D 0.000e+00 0.000e+00 NaN NaN\nmodelMerc 280 3.537e-08 3.387e+06 0 1\nmodelMerc 280C -1.763e-05 3.113e+06 0 1\nmodelMerc 450SE 4.560e-10 8.472e+06 0 1\nmodelMerc 450SL 2.063e-10 3.993e+06 0 1\nmodelMerc 450SLC 2.581e-10 4.887e+06 0 1\nmodelPontiac Firebird 2.934e-10 5.495e+06 0 1\nmodelPorsche 914-2 -5.713e+01 8.485e+06 0 1\nmodelToyota Corolla -1.764e-05 3.047e+06 0 1\nmodelToyota Corona 5.713e+01 1.270e+07 0 1\nmodelValiant 5.713e+01 1.128e+06 0 1\nmodelVolvo 142E 8.495e-07 1.543e+07 0 1\namTRUE -6.645e-09 2.111e+07 0 1\ngear.L 4.040e+01 3.334e+06 0 1\ngear.Q -2.332e+01 7.942e+05 0 1\ncarb 3.703e-12 4.368e+05 0 1\n\nApproximate significance of smooth terms:\n edf Ref.df Chi.sq p-value\ns(wt) 1 1 0 1\n\nRank: 42/45\nR-sq.(adj) = 1 Deviance explained = 100%\nUBRE = 0.0625 Scale est. = 1 n = 64"
attr(,"params")$y_col
[1] "vs"
attr(,"params")$rand_it
[1] 10
attr(,"params")$parallel
[1] 0
attr(,"params")$model_packages
NULL
attr(,"params")$random_model_call_string
NULL
attr(,"params")$random_model_call_string_vars
character(0)
attr(,"params")$positive
[1] TRUE
attr(,"params")$seed
[1] 0
attr(,"params")$rand_it_ok
[1] 10
attr(,"params")$exactness
[1] "invalid"
ALEpDist works with categorical outcome
Code
unclass(pd)
Output
<object>
attr(,"S7_class")
<ale::ALEpDist> class
@ parent : <S7_object>
@ constructor: function(model, data, ..., y_col, rand_it, surrogate, parallel, model_packages, random_model_call_string, random_model_call_string_vars, positive, pred_fun, pred_type, output_residuals, seed, silent, .skip_validation) {...}
@ validator : <NULL>
@ properties :
$ rand_stats : <list>
$ residual_distribution: S3<univariateML>
$ residuals : <double> or <NULL>
$ params : <list>
attr(,"rand_stats")
attr(,"rand_stats")$Asia
# A tibble: 10 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 0 0 0 0 0 0
2 0 0 0 0 0 0
3 0 0 0 0 0 0
4 0 0 0 0 0 0
5 0 0 0 0 0 0
6 0 0 0 0 0 0
7 0 0 0 0 0 0
8 0 0 0 0 0 0
9 0 0 0 0 0 0
10 0 0 0 0 0 0
attr(,"rand_stats")$Europe
# A tibble: 10 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 0 0 0 0 0 0
2 0 0 0 0 0 0
3 0 0 0 0 0 0
4 0 0 0 0 0 0
5 0 0 0 0 0 0
6 0 0 0 0 0 0
7 0 0 0 0 0 0
8 0 0 0 0 0 0
9 0 0 0 0 0 0
10 0 0 0 0 0 0
attr(,"rand_stats")$`North America`
# A tibble: 10 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 0 0 0 0 0 0
2 0 0 0 0 0 0
3 0 0 0 0 0 0
4 0 0 0 0 0 0
5 0 0 0 0 0 0
6 0 0 0 0 0 0
7 0 0 0 0 0 0
8 0 0 0 0 0 0
9 0 0 0 0 0 0
10 0 0 0 0 0 0
attr(,"residual_distribution")
Maximum likelihood estimates for the Laplace model
mu sigma
-2.043e-23 1.503e-17
attr(,"params")
attr(,"params")$model
attr(,"params")$model$class
[1] "multinom" "nnet"
attr(,"params")$model$call
[1] "nnet::multinom(formula = continent ~ model + wt + am + gear + \n carb, data = test_cars, trace = FALSE)"
attr(,"params")$model$print
[1] "Call:\nnnet::multinom(formula = continent ~ model + wt + am + gear + \n carb, data = test_cars, trace = FALSE)\n\nCoefficients:\n (Intercept) modelCadillac Fleetwood modelCamaro Z28\nEurope -21.86205 -2.222771 -6.526356\nNorth America -12.29302 -2.006450 14.817906\n modelChrysler Imperial modelDatsun 710 modelDodge Challenger\nEurope -3.5495185 -42.28844 -3.766810\nNorth America -0.4119232 -22.35666 7.689393\n modelDuster 360 modelFerrari Dino modelFiat 128 modelFiat X1-9\nEurope -7.046867 5.545836 42.144872 43.426334\nNorth America 20.851301 -3.241914 2.522922 2.870462\n modelFord Pantera L modelHonda Civic modelHornet 4 Drive\nEurope -23.11127 -27.5878163 -2.495884\nNorth America 29.30682 -0.1513286 7.771946\n modelHornet Sportabout modelLincoln Continental modelLotus Europa\nEurope -5.036042 -1.780289 14.561696\nNorth America 12.103127 -2.891703 -1.631007\n modelMaserati Bora modelMazda RX4 modelMazda RX4 Wag\nEurope 6.999218 -33.61112 -35.88135\nNorth America -7.857123 -3.27005 -10.72975\n modelMerc 230 modelMerc 240D modelMerc 280 modelMerc 280C\nEurope 14.2684900 14.2452958 11.590028 10.4496565\nNorth America -0.5204999 -0.6708756 1.228568 0.8928993\n modelMerc 450SE modelMerc 450SL modelMerc 450SLC\nEurope 39.49499 34.5616 36.09627\nNorth America -38.59611 -31.8537 -34.17067\n modelPontiac Firebird modelPorsche 914-2 modelToyota Corolla\nEurope -3.057543 15.344405 -34.531039\nNorth America 4.399831 -4.102117 -8.804816\n modelToyota Corona modelValiant modelVolvo 142E wt\nEurope -44.72605 -2.073514 34.932615 4.534617\nNorth America -52.77315 5.537827 2.626442 3.625818\n amTRUE gear.L gear.Q carb\nEurope -34.05607 40.7995829 -5.443825 10.524876\nNorth America -24.81811 0.6220113 39.516400 5.894396\n\nResidual Deviance: 0.0001161393 \nAIC: 136.0001 "
attr(,"params")$model$summary
[1] "Call:\nnnet::multinom(formula = continent ~ model + wt + am + gear + \n carb, data = test_cars, trace = FALSE)\n\nCoefficients:\n (Intercept) modelCadillac Fleetwood modelCamaro Z28\nEurope -21.86205 -2.222771 -6.526356\nNorth America -12.29302 -2.006450 14.817906\n modelChrysler Imperial modelDatsun 710 modelDodge Challenger\nEurope -3.5495185 -42.28844 -3.766810\nNorth America -0.4119232 -22.35666 7.689393\n modelDuster 360 modelFerrari Dino modelFiat 128 modelFiat X1-9\nEurope -7.046867 5.545836 42.144872 43.426334\nNorth America 20.851301 -3.241914 2.522922 2.870462\n modelFord Pantera L modelHonda Civic modelHornet 4 Drive\nEurope -23.11127 -27.5878163 -2.495884\nNorth America 29.30682 -0.1513286 7.771946\n modelHornet Sportabout modelLincoln Continental modelLotus Europa\nEurope -5.036042 -1.780289 14.561696\nNorth America 12.103127 -2.891703 -1.631007\n modelMaserati Bora modelMazda RX4 modelMazda RX4 Wag\nEurope 6.999218 -33.61112 -35.88135\nNorth America -7.857123 -3.27005 -10.72975\n modelMerc 230 modelMerc 240D modelMerc 280 modelMerc 280C\nEurope 14.2684900 14.2452958 11.590028 10.4496565\nNorth America -0.5204999 -0.6708756 1.228568 0.8928993\n modelMerc 450SE modelMerc 450SL modelMerc 450SLC\nEurope 39.49499 34.5616 36.09627\nNorth America -38.59611 -31.8537 -34.17067\n modelPontiac Firebird modelPorsche 914-2 modelToyota Corolla\nEurope -3.057543 15.344405 -34.531039\nNorth America 4.399831 -4.102117 -8.804816\n modelToyota Corona modelValiant modelVolvo 142E wt\nEurope -44.72605 -2.073514 34.932615 4.534617\nNorth America -52.77315 5.537827 2.626442 3.625818\n amTRUE gear.L gear.Q carb\nEurope -34.05607 40.7995829 -5.443825 10.524876\nNorth America -24.81811 0.6220113 39.516400 5.894396\n\nStd. Errors:\n (Intercept) modelCadillac Fleetwood modelCamaro Z28\nEurope 35852.83 64.65729 NaN\nNorth America 12968.78 64.70730 2.870067e-08\n modelChrysler Imperial modelDatsun 710 modelDodge Challenger\nEurope 477.3874 1.590757e-08 2.858243e-08\nNorth America 477.3874 2.200012e-18 6.428719e+00\n modelDuster 360 modelFerrari Dino modelFiat 128 modelFiat X1-9\nEurope 5.486580e-10 2660.898 1.805399e+04 1.525317e+04\nNorth America 7.249665e-11 2660.898 1.192455e-12 4.867466e-13\n modelFord Pantera L modelHonda Civic modelHornet 4 Drive\nEurope 0.09603878 5.871595e-07 3.863891e-10\nNorth America 0.09624948 9.331882e-08 7.163294e+03\n modelHornet Sportabout modelLincoln Continental modelLotus Europa\nEurope 4.655689e-09 285.0759 32194.07\nNorth America 5.351209e-02 285.1436 20099.45\n modelMaserati Bora modelMazda RX4 modelMazda RX4 Wag\nEurope 0.01739745 644.72277405 1.000913e+02\nNorth America 0.01739745 0.04282998 3.394594e-05\n modelMerc 230 modelMerc 240D modelMerc 280 modelMerc 280C\nEurope 1.329222e+02 1.088958e+02 4.023889e-07 2.701645e-06\nNorth America 4.475756e-08 3.800080e-08 2.855911e-10 1.280090e-09\n modelMerc 450SE modelMerc 450SL modelMerc 450SLC\nEurope 92448.24826 22918.50 1179.3388\nNorth America 12.05893 21677.09 776.0876\n modelPontiac Firebird modelPorsche 914-2 modelToyota Corolla\nEurope 1.504060e-06 29024.38 9.393802e-15\nNorth America 6.422677e+01 62881.87 2.008891e-13\n modelToyota Corona modelValiant modelVolvo 142E wt amTRUE\nEurope 2.042457e-17 7.332842e-09 5.751337e+04 26852.73 81773.79\nNorth America 1.689525e+01 3.329289e+04 5.749185e-12 97682.05 57698.40\n gear.L gear.Q carb\nEurope 120502.95 45794.583 77388.60\nNorth America 84695.32 5294.488 34719.66\n\nResidual Deviance: 0.0001161393 \nAIC: 136.0001 "
attr(,"params")$y_col
[1] "continent"
attr(,"params")$rand_it
[1] 10
attr(,"params")$parallel
[1] 0
attr(,"params")$model_packages
[1] "nnet"
attr(,"params")$random_model_call_string
NULL
attr(,"params")$random_model_call_string_vars
character(0)
attr(,"params")$positive
[1] TRUE
attr(,"params")$seed
[1] 0
attr(,"params")$rand_it_ok
[1] 10
attr(,"params")$exactness
[1] "invalid"
Try the ale package in your browser
Any scripts or data that you put into this service are public.
ale documentation built on April 11, 2025, 6:09 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
ALEpDist works with default inputs (exact) on ALE()
Code
unclass(pd)
Output
<object>
attr(,"S7_class")
<ale::ALEpDist> class
@ parent : <S7_object>
@ constructor: function(model, data, ..., y_col, rand_it, surrogate, parallel, model_packages, random_model_call_string, random_model_call_string_vars, positive, pred_fun, pred_type, output_residuals, seed, silent, .skip_validation) {...}
@ validator : <NULL>
@ properties :
$ rand_stats : <list>
$ residual_distribution: S3<univariateML>
$ residuals : <double> or <NULL>
$ params : <list>
attr(,"rand_stats")
attr(,"rand_stats")$mpg
# A tibble: 10 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 0.000484 -0.00330 0.00201 0 0 0
2 0.00211 -0.00659 0.00835 0.342 -1.56 1.56
3 0.00196 -0.00644 0.00866 0.220 -1.56 1.56
4 0.000908 -0.00363 0.00524 0.0244 -1.56 1.56
5 0.000352 -0.00166 0.00139 0 0 0
6 0.000389 -0.00192 0.00158 0 0 0
7 0.00136 -0.00551 0.00407 0.0977 -1.56 1.56
8 0.000976 -0.00361 0.00715 0.0488 -1.56 1.56
9 0.00280 -0.0136 0.00822 0.439 -1.56 1.56
10 0.000472 -0.00171 0.00149 0 0 0
attr(,"residual_distribution")
Maximum likelihood estimates for the Laplace model
mu sigma
1.303e-11 3.587e-03
attr(,"params")
attr(,"params")$model
attr(,"params")$model$class
[1] "gam" "glm" "lm"
attr(,"params")$model$call
[1] "mgcv::gam(formula = mpg ~ model + s(wt) + am + gear + carb, data = test_cars)"
attr(,"params")$model$print
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nEstimated degrees of freedom:\n8.03 total = 41.03 \n\nGCV score: 0.0001770391 rank: 42/45"
attr(,"params")$model$summary
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nParametric coefficients:\n Estimate Std. Error t value Pr(>|t|) \n(Intercept) 1.432e+01 1.353e-01 105.784 < 2e-16 ***\nmodelCadillac Fleetwood -9.910e+00 1.259e+00 -7.873 5.68e-08 ***\nmodelCamaro Z28 -3.700e+00 7.268e-02 -50.911 < 2e-16 ***\nmodelChrysler Imperial -5.777e+00 1.276e+00 -4.526 0.000152 ***\nmodelDatsun 710 -3.793e+00 1.131e-01 -33.550 < 2e-16 ***\nmodelDodge Challenger -1.266e-01 2.060e-02 -6.147 2.87e-06 ***\nmodelDuster 360 -1.547e+00 2.851e-02 -54.276 < 2e-16 ***\nmodelFerrari Dino -4.088e+00 1.542e-01 -26.506 < 2e-16 ***\nmodelFiat 128 7.211e+00 9.518e-02 75.763 < 2e-16 ***\nmodelFiat X1-9 5.916e+00 1.941e-01 30.488 < 2e-16 ***\nmodelFord Pantera L -1.094e+01 1.737e-01 -63.000 < 2e-16 ***\nmodelHonda Civic 1.474e+01 2.896e-01 50.893 < 2e-16 ***\nmodelHornet 4 Drive 7.569e+00 5.315e-02 142.406 < 2e-16 ***\nmodelHornet Sportabout 3.468e+00 9.616e-03 360.698 < 2e-16 ***\nmodelLincoln Continental -1.023e+01 1.279e+00 -7.998 4.34e-08 ***\nmodelLotus Europa 2.341e+01 3.392e-01 69.015 < 2e-16 ***\nmodelMaserati Bora -1.408e+01 1.903e-01 -74.006 < 2e-16 ***\nmodelMazda RX4 -8.359e+00 1.638e-01 -51.017 < 2e-16 ***\nmodelMazda RX4 Wag -1.030e+01 1.761e-01 -58.494 < 2e-16 ***\nmodelMerc 230 2.481e+00 5.506e-02 45.064 < 2e-16 ***\nmodelMerc 240D 3.804e+00 5.586e-02 68.099 < 2e-16 ***\nmodelMerc 280 -2.984e+00 6.794e-02 -43.926 < 2e-16 ***\nmodelMerc 280C -4.382e+00 6.668e-02 -65.723 < 2e-16 ***\nmodelMerc 450SE -1.661e+00 1.075e-01 -15.448 1.26e-13 ***\nmodelMerc 450SL 7.892e-01 5.311e-02 14.861 2.83e-13 ***\nmodelMerc 450SLC -1.524e+00 6.416e-02 -23.749 < 2e-16 ***\nmodelPontiac Firebird 2.178e+00 7.002e-02 31.102 < 2e-16 ***\nmodelPorsche 914-2 8.306e+00 1.409e-01 58.945 < 2e-16 ***\nmodelToyota Corolla 1.419e+01 2.372e-01 59.809 < 2e-16 ***\nmodelToyota Corona 1.342e+01 2.208e-01 60.795 < 2e-16 ***\nmodelValiant 2.760e+00 1.050e-02 262.897 < 2e-16 ***\nmodelVolvo 142E -9.189e+00 1.720e-01 -53.428 < 2e-16 ***\namTRUE 1.302e+01 1.792e-01 72.629 < 2e-16 ***\ngear.L 1.571e-01 2.703e-02 5.811 6.42e-06 ***\ngear.Q -5.584e+00 4.818e-02 -115.914 < 2e-16 ***\ncarb -3.135e-04 4.119e-03 -0.076 0.939977 \n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nApproximate significance of smooth terms:\n edf Ref.df F p-value \ns(wt) 8.027 8.693 449 <2e-16 ***\n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nRank: 42/45\nR-sq.(adj) = 1 Deviance explained = 100%\nGCV = 0.00017704 Scale est. = 6.3549e-05 n = 64"
attr(,"params")$y_col
[1] "mpg"
attr(,"params")$rand_it
[1] 10
attr(,"params")$parallel
[1] 0
attr(,"params")$model_packages
NULL
attr(,"params")$random_model_call_string
NULL
attr(,"params")$random_model_call_string_vars
character(0)
attr(,"params")$positive
[1] TRUE
attr(,"params")$seed
[1] 0
attr(,"params")$rand_it_ok
[1] 10
attr(,"params")$exactness
[1] "invalid"
Code
unclass(cars_ale)
Output
<object>
attr(,"S7_class")
<ale::ALE> class
@ parent : <S7_object>
@ constructor: function(model, x_cols, data, y_col, ..., exclude_cols, parallel, model_packages, output_stats, output_boot_data, pred_fun, pred_type, p_values, aler_alpha, max_num_bins, boot_it, boot_alpha, boot_centre, seed, y_type, sample_size, silent, .bins) {...}
@ validator : <NULL>
@ properties :
$ effect: <list>
$ params: <list>
attr(,"effect")
attr(,"effect")$mpg
attr(,"effect")$mpg$ale
attr(,"effect")$mpg$ale$d1
attr(,"effect")$mpg$ale$d1$vs
# A tibble: 2 x 7
vs.bin .n .y .y_lo .y_mean .y_median .y_hi
<ord> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 FALSE 36 0 0 0 0 0
2 TRUE 28 0 0 0 0 0
attr(,"effect")$mpg$ale$d1$continent
# A tibble: 3 x 7
continent.bin .n .y .y_lo .y_mean .y_median .y_hi
<ord> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 North America 24 0 0 0 0 0
2 Europe 28 0 0 0 0 0
3 Asia 12 0 0 0 0 0
attr(,"effect")$mpg$ale$d1$am
# A tibble: 2 x 7
am.bin .n .y .y_lo .y_mean .y_median .y_hi
<ord> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 FALSE 38 -1.61 -5.01 -1.61 -1.01 0.775
2 TRUE 26 1.60 -1.75 1.60 0.219 7.28
attr(,"effect")$mpg$ale$d1$model
# A tibble: 32 x 7
model.bin .n .y .y_lo .y_mean .y_median .y_hi
<ord> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 Camaro Z28 2 -1.22 -3.83 -1.22 -1.74 2.27
2 Cadillac Fleetwood 2 -0.330 -9.42 -0.330 0.233 7.80
3 Lincoln Continental 2 7.05 -9.18 7.05 7.26 22.9
4 Chrysler Imperial 2 15.5 -4.52 15.5 14.3 37.4
5 Duster 360 2 21.7 -0.182 21.7 21.6 43.9
6 Hornet Sportabout 2 31.8 5.16 31.8 37.6 48.4
7 Pontiac Firebird 2 25.1 3.36 25.1 29.3 39.7
8 Dodge Challenger 2 20.3 0.910 20.3 21.6 37.4
9 AMC Javelin 2 20.4 1.16 20.4 22.0 36.9
10 Merc 450SL 2 20.8 1.13 20.8 15.1 50.2
# i 22 more rows
attr(,"effect")$mpg$ale$d1$gear
# A tibble: 3 x 7
gear.bin .n .y .y_lo .y_mean .y_median .y_hi
<ord> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 three 30 -1.06 -2.54 -1.06 -0.758 -0.0935
2 four 24 2.66 1.49 2.66 2.42 4.22
3 five 10 -1.60 -2.46 -1.60 -1.85 -0.331
attr(,"effect")$mpg$ale$d1$carb
# A tibble: 5 x 7
carb.ceil .n .y .y_lo .y_mean .y_median .y_hi
<dbl> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 1 14 0.000490 0.000490 0.000490 0.000490 0.000490
2 2 19 0.000176 0.000176 0.000176 0.000176 0.000176
3 3 9 -0.000137 -0.000137 -0.000137 -0.000137 -0.000137
4 4 16 -0.000451 -0.000451 -0.000451 -0.000451 -0.000451
5 8 6 -0.00170 -0.00170 -0.00170 -0.00170 -0.00170
attr(,"effect")$mpg$ale$d1$wt
# A tibble: 10 x 7
wt.ceil .n .y .y_lo .y_mean .y_median .y_hi
<dbl> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 1.50 1 -18.1 -18.1 -18.1 -18.1 -18.1
2 1.94 7 -10.2 -10.2 -10.2 -10.2 -10.2
3 2.46 7 -3.51 -3.51 -3.51 -3.51 -3.51
4 2.79 7 -0.882 -0.882 -0.882 -0.882 -0.882
5 3.19 7 2.07 2.07 2.07 2.07 2.07
6 3.44 7 3.65 3.65 3.65 3.65 3.65
7 3.52 7 4.05 4.05 4.05 4.05 4.05
8 3.73 7 4.92 4.92 4.92 4.92 4.92
9 4.05 7 6.38 6.38 6.38 6.38 6.38
10 5.45 7 9.11 9.11 9.11 9.11 9.11
attr(,"effect")$mpg$stats
attr(,"effect")$mpg$stats$d1
# A tibble: 42 x 8
statistic estimate p.value term conf.low mean median conf.high
<chr> <dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 aled 0 1 vs 0 0 0 0
2 aler_min 0 1 vs 0 0 0 0
3 aler_max 0 1 vs 0 0 0 0
4 naled 0 0.6 vs 0 0 0 0
5 naler_min 0 1 vs 0 0 0 0
6 naler_max 0 0.6 vs 0 0 0 0
7 aled 0 1 continent 0 0 0 0
8 aler_min 0 1 continent 0 0 0 0
9 aler_max 0 1 continent 0 0 0 0
10 naled 0 0.6 continent 0 0 0 0
# i 32 more rows
attr(,"effect")$mpg$boot_data
NULL
attr(,"params")
attr(,"params")$max_d
[1] 1
attr(,"params")$ordered_x_cols
attr(,"params")$ordered_x_cols$d1
[1] "vs" "continent" "am" "model" "gear" "carb"
[7] "wt"
attr(,"params")$ordered_x_cols$d2
character(0)
attr(,"params")$requested_x_cols
attr(,"params")$requested_x_cols$d1
[1] "vs" "continent" "am" "model" "gear" "carb"
[7] "wt"
attr(,"params")$requested_x_cols$d2
character(0)
attr(,"params")$y_cats
[1] "mpg"
attr(,"params")$y_summary
mpg
min 10.39108
1% 10.39108
2.5% 10.40000
5% 10.88271
10% 14.33418
20% 15.16500
25% 15.43921
30% 15.79628
40% 17.83840
aler_lo_lo 19.18669
aler_lo 19.18797
50% 19.20000
mean 20.09462
aler_hi 19.20859
aler_hi_hi 19.20864
60% 21.00000
70% 21.51193
75% 22.80000
80% 24.48680
90% 30.31124
95% 32.14486
97.5% 33.08402
99% 33.84876
max 33.84876
attr(,"params")$model
attr(,"params")$model$class
[1] "gam" "glm" "lm"
attr(,"params")$model$call
[1] "mgcv::gam(formula = mpg ~ model + s(wt) + am + gear + carb, data = test_cars)"
attr(,"params")$model$print
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nEstimated degrees of freedom:\n8.03 total = 41.03 \n\nGCV score: 0.0001770391 rank: 42/45"
attr(,"params")$model$summary
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nParametric coefficients:\n Estimate Std. Error t value Pr(>|t|) \n(Intercept) 1.432e+01 1.353e-01 105.784 < 2e-16 ***\nmodelCadillac Fleetwood -9.910e+00 1.259e+00 -7.873 5.68e-08 ***\nmodelCamaro Z28 -3.700e+00 7.268e-02 -50.911 < 2e-16 ***\nmodelChrysler Imperial -5.777e+00 1.276e+00 -4.526 0.000152 ***\nmodelDatsun 710 -3.793e+00 1.131e-01 -33.550 < 2e-16 ***\nmodelDodge Challenger -1.266e-01 2.060e-02 -6.147 2.87e-06 ***\nmodelDuster 360 -1.547e+00 2.851e-02 -54.276 < 2e-16 ***\nmodelFerrari Dino -4.088e+00 1.542e-01 -26.506 < 2e-16 ***\nmodelFiat 128 7.211e+00 9.518e-02 75.763 < 2e-16 ***\nmodelFiat X1-9 5.916e+00 1.941e-01 30.488 < 2e-16 ***\nmodelFord Pantera L -1.094e+01 1.737e-01 -63.000 < 2e-16 ***\nmodelHonda Civic 1.474e+01 2.896e-01 50.893 < 2e-16 ***\nmodelHornet 4 Drive 7.569e+00 5.315e-02 142.406 < 2e-16 ***\nmodelHornet Sportabout 3.468e+00 9.616e-03 360.698 < 2e-16 ***\nmodelLincoln Continental -1.023e+01 1.279e+00 -7.998 4.34e-08 ***\nmodelLotus Europa 2.341e+01 3.392e-01 69.015 < 2e-16 ***\nmodelMaserati Bora -1.408e+01 1.903e-01 -74.006 < 2e-16 ***\nmodelMazda RX4 -8.359e+00 1.638e-01 -51.017 < 2e-16 ***\nmodelMazda RX4 Wag -1.030e+01 1.761e-01 -58.494 < 2e-16 ***\nmodelMerc 230 2.481e+00 5.506e-02 45.064 < 2e-16 ***\nmodelMerc 240D 3.804e+00 5.586e-02 68.099 < 2e-16 ***\nmodelMerc 280 -2.984e+00 6.794e-02 -43.926 < 2e-16 ***\nmodelMerc 280C -4.382e+00 6.668e-02 -65.723 < 2e-16 ***\nmodelMerc 450SE -1.661e+00 1.075e-01 -15.448 1.26e-13 ***\nmodelMerc 450SL 7.892e-01 5.311e-02 14.861 2.83e-13 ***\nmodelMerc 450SLC -1.524e+00 6.416e-02 -23.749 < 2e-16 ***\nmodelPontiac Firebird 2.178e+00 7.002e-02 31.102 < 2e-16 ***\nmodelPorsche 914-2 8.306e+00 1.409e-01 58.945 < 2e-16 ***\nmodelToyota Corolla 1.419e+01 2.372e-01 59.809 < 2e-16 ***\nmodelToyota Corona 1.342e+01 2.208e-01 60.795 < 2e-16 ***\nmodelValiant 2.760e+00 1.050e-02 262.897 < 2e-16 ***\nmodelVolvo 142E -9.189e+00 1.720e-01 -53.428 < 2e-16 ***\namTRUE 1.302e+01 1.792e-01 72.629 < 2e-16 ***\ngear.L 1.571e-01 2.703e-02 5.811 6.42e-06 ***\ngear.Q -5.584e+00 4.818e-02 -115.914 < 2e-16 ***\ncarb -3.135e-04 4.119e-03 -0.076 0.939977 \n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nApproximate significance of smooth terms:\n edf Ref.df F p-value \ns(wt) 8.027 8.693 449 <2e-16 ***\n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nRank: 42/45\nR-sq.(adj) = 1 Deviance explained = 100%\nGCV = 0.00017704 Scale est. = 6.3549e-05 n = 64"
attr(,"params")$data
attr(,"params")$data$data_sample
# A tibble: 64 x 8
mpg vs continent am model gear carb wt
<dbl> <lgl> <fct> <lgl> <chr> <ord> <int> <dbl>
1 21 FALSE Asia TRUE Mazda RX4 four 4 2.62
2 21 FALSE Asia TRUE Mazda RX4 Wag four 4 2.88
3 22.8 TRUE Asia TRUE Datsun 710 four 1 2.32
4 21.4 TRUE North America FALSE Hornet 4 Drive three 1 3.22
5 18.7 FALSE North America FALSE Hornet Sportabout three 2 3.44
6 18.1 TRUE North America FALSE Valiant three 1 3.46
7 14.3 FALSE North America FALSE Duster 360 three 4 3.57
8 24.4 TRUE Europe FALSE Merc 240D four 2 3.19
9 22.8 TRUE Europe FALSE Merc 230 four 2 3.15
10 19.2 TRUE Europe FALSE Merc 280 four 4 3.44
# i 54 more rows
attr(,"params")$data$y_vals_sample
mpg
[1,] 21.00000
[2,] 21.00000
[3,] 22.80000
[4,] 21.40000
[5,] 18.70000
[6,] 18.10000
[7,] 14.30000
[8,] 24.40000
[9,] 22.80000
[10,] 19.20000
[11,] 17.80000
[12,] 16.40000
[13,] 17.30000
[14,] 15.20000
[15,] 10.40000
[16,] 10.40000
[17,] 14.70000
[18,] 32.40000
[19,] 30.40000
[20,] 33.90000
[21,] 21.50000
[22,] 15.50000
[23,] 15.20000
[24,] 13.30000
[25,] 19.20000
[26,] 27.30000
[27,] 26.00000
[28,] 30.40000
[29,] 15.80000
[30,] 19.70000
[31,] 15.00000
[32,] 21.40000
[33,] 21.16661
[34,] 20.90151
[35,] 22.74169
[36,] 21.43118
[37,] 18.85267
[38,] 17.99201
[39,] 14.41394
[40,] 24.61700
[41,] 22.87332
[42,] 19.24958
[43,] 17.64400
[44,] 16.30356
[45,] 17.18809
[46,] 15.25685
[47,] 10.37589
[48,] 10.45613
[49,] 14.69932
[50,] 32.54102
[51,] 30.69908
[52,] 33.81866
[53,] 21.61930
[54,] 15.63476
[55,] 15.11249
[56,] 13.34035
[57,] 19.05621
[58,] 27.17290
[59,] 25.94078
[60,] 30.10414
[61,] 15.76283
[62,] 19.84566
[63,] 14.95210
[64,] 21.39233
attr(,"params")$data$nrow
[1] 64
attr(,"params")$y_col
[1] "mpg"
attr(,"params")$parallel
[1] 0
attr(,"params")$model_packages
NULL
attr(,"params")$output_stats
[1] TRUE
attr(,"params")$output_boot_data
[1] FALSE
attr(,"params")$pred_fun
[1] "function(object, newdata, type = pred_type) {\n stats::predict(object = object, newdata = newdata, type = type)\n }"
attr(,"params")$pred_type
[1] "response"
attr(,"params")$p_values
<ale::ALEpDist>
@ rand_stats :List of 1
.. $ mpg: tibble [10 x 6] (S3: tbl_df/tbl/data.frame)
.. ..$ aled : num [1:10] 0.000484 0.002108 0.001961 0.000908 0.000352 ...
.. ..$ aler_min : num [1:10] -0.0033 -0.00659 -0.00644 -0.00363 -0.00166 ...
.. ..$ aler_max : num [1:10] 0.00201 0.00835 0.00866 0.00524 0.00139 ...
.. ..$ naled : num [1:10] 0 0.3418 0.2197 0.0244 0 ...
.. ..$ naler_min: num [1:10] 0 -1.56 -1.56 -1.56 0 ...
.. ..$ naler_max: num [1:10] 0 1.56 1.56 1.56 0 ...
@ residual_distribution: 'univariateML' Named num [1:2] 1.30e-11 3.59e-03
.. - attr(*, "logLik")= num 252
.. - attr(*, "call")= language f(x = x, na.rm = na.rm)
.. - attr(*, "n")= int 64
.. - attr(*, "model")= chr "Laplace"
.. - attr(*, "density")= chr "extraDistr::dlaplace"
.. - attr(*, "support")= num [1:2] -Inf Inf
.. - attr(*, "names")= chr [1:2] "mu" "sigma"
.. - attr(*, "default")= num [1:2] 0 1
.. - attr(*, "continuous")= logi TRUE
@ residuals : NULL
@ params :List of 11
.. $ model :List of 4
.. ..$ class : chr [1:3] "gam" "glm" "lm"
.. ..$ call : chr "mgcv::gam(formula = mpg ~ model + s(wt) + am + gear + carb, data = test_cars)"
.. ..$ print : chr "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nEstimated "| __truncated__
.. ..$ summary: chr "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nParametric"| __truncated__
.. $ y_col : chr "mpg"
.. $ rand_it : num 10
.. $ parallel : num 0
.. $ model_packages : NULL
.. $ random_model_call_string : NULL
.. $ random_model_call_string_vars: chr(0)
.. $ positive : logi TRUE
.. $ seed : num 0
.. $ rand_it_ok : int 10
.. $ exactness : chr "invalid"
attr(,"params")$aler_alpha
[1] 0.01 0.05
attr(,"params")$max_num_bins
[1] 10
attr(,"params")$boot_it
[1] 3
attr(,"params")$boot_alpha
[1] 0.05
attr(,"params")$boot_centre
[1] "mean"
attr(,"params")$seed
[1] 0
attr(,"params")$y_type
[1] "numeric"
attr(,"params")$sample_size
[1] 500
Code
unclass(set_names(map(stats_names, function(.stat) {
value_to_p(pd@rand_stats$mpg, .stat, test_vals)
}), stats_names))
Output
$aled
[1] 1 1 1 1 1 0 0 0 0 0 0
$aler_min
[1] 0 0 0 0 1 1 1 1 1 1 1
$aler_max
[1] 1 1 1 1 1 0 0 0 0 0 0
$naled
[1] 1.0 1.0 1.0 1.0 0.6 0.4 0.3 0.0 0.0 0.0 0.0
$naler_min
[1] 0.0 0.0 0.6 0.6 1.0 1.0 1.0 1.0 1.0 1.0 1.0
$naler_max
[1] 1.0 1.0 1.0 1.0 0.6 0.6 0.6 0.6 0.6 0.0 0.0
Surrogate ALEpDist works
Code
unclass(pd)
Output
<object>
attr(,"S7_class")
<ale::ALEpDist> class
@ parent : <S7_object>
@ constructor: function(model, data, ..., y_col, rand_it, surrogate, parallel, model_packages, random_model_call_string, random_model_call_string_vars, positive, pred_fun, pred_type, output_residuals, seed, silent, .skip_validation) {...}
@ validator : <NULL>
@ properties :
$ rand_stats : <list>
$ residual_distribution: S3<univariateML>
$ residuals : <double> or <NULL>
$ params : <list>
attr(,"rand_stats")
attr(,"rand_stats")$mpg
# A tibble: 100 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 0.00199 -0.0112 0.00727 0.366 -1.56 1.56
2 0.00394 -0.0118 0.0148 0.708 -1.56 1.56
3 0.00425 -0.0170 0.0126 0.708 -1.56 1.56
4 0.00647 -0.0215 0.0304 0.708 -1.56 1.56
5 0.00365 -0.0149 0.0124 0.537 -1.56 1.56
6 0.00281 -0.0121 0.0100 0.366 -1.56 1.56
7 0.00599 -0.0172 0.0227 0.879 -1.56 1.56
8 0.00574 -0.0333 0.0175 0.708 -1.56 1.56
9 0.00733 -0.0321 0.0203 0.879 -1.56 1.56
10 0.00729 -0.0228 0.0257 1.05 -1.56 1.56
# i 90 more rows
attr(,"residual_distribution")
Maximum likelihood estimates for the Laplace model
mu sigma
1.303e-11 3.587e-03
attr(,"residuals")
[1] -9.470698e-04 -1.130145e-03 -3.078035e-03 7.415332e-04 -4.678952e-03
[6] 7.516518e-04 2.728091e-03 -8.853029e-03 -3.016706e-04 -1.794893e-03
[11] -3.673897e-03 -2.816578e-03 5.414042e-03 2.979146e-03 2.219206e-03
[16] 8.324011e-04 -4.976941e-05 -1.115543e-02 1.437735e-03 2.200997e-03
[21] 2.625747e-03 5.178720e-04 -9.802341e-03 7.118944e-03 5.255702e-03
[26] -9.746617e-03 -2.976337e-03 6.542735e-03 -8.071930e-03 4.016990e-03
[31] -2.747836e-04 -6.032700e-05 9.470698e-04 1.130145e-03 3.078035e-03
[36] -7.415332e-04 4.678952e-03 -7.516518e-04 -2.728091e-03 8.853029e-03
[41] 3.016706e-04 1.794893e-03 3.673897e-03 2.816578e-03 -5.414042e-03
[46] -2.979146e-03 -2.219206e-03 -8.324011e-04 4.976944e-05 1.115543e-02
[51] -1.437735e-03 -2.200997e-03 -2.625747e-03 -5.178720e-04 9.802341e-03
[56] -7.118944e-03 -5.255702e-03 9.746617e-03 2.976337e-03 -6.542735e-03
[61] 8.071930e-03 -4.016990e-03 2.747836e-04 6.032702e-05
attr(,"params")
attr(,"params")$model
attr(,"params")$model$class
[1] "gam" "glm" "lm"
attr(,"params")$model$call
[1] "mgcv::gam(formula = mpg ~ model + s(wt) + am + gear + carb, data = test_cars)"
attr(,"params")$model$print
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nEstimated degrees of freedom:\n8.03 total = 41.03 \n\nGCV score: 0.0001770391 rank: 42/45"
attr(,"params")$model$summary
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nParametric coefficients:\n Estimate Std. Error t value Pr(>|t|) \n(Intercept) 1.432e+01 1.353e-01 105.784 < 2e-16 ***\nmodelCadillac Fleetwood -9.910e+00 1.259e+00 -7.873 5.68e-08 ***\nmodelCamaro Z28 -3.700e+00 7.268e-02 -50.911 < 2e-16 ***\nmodelChrysler Imperial -5.777e+00 1.276e+00 -4.526 0.000152 ***\nmodelDatsun 710 -3.793e+00 1.131e-01 -33.550 < 2e-16 ***\nmodelDodge Challenger -1.266e-01 2.060e-02 -6.147 2.87e-06 ***\nmodelDuster 360 -1.547e+00 2.851e-02 -54.276 < 2e-16 ***\nmodelFerrari Dino -4.088e+00 1.542e-01 -26.506 < 2e-16 ***\nmodelFiat 128 7.211e+00 9.518e-02 75.763 < 2e-16 ***\nmodelFiat X1-9 5.916e+00 1.941e-01 30.488 < 2e-16 ***\nmodelFord Pantera L -1.094e+01 1.737e-01 -63.000 < 2e-16 ***\nmodelHonda Civic 1.474e+01 2.896e-01 50.893 < 2e-16 ***\nmodelHornet 4 Drive 7.569e+00 5.315e-02 142.406 < 2e-16 ***\nmodelHornet Sportabout 3.468e+00 9.616e-03 360.698 < 2e-16 ***\nmodelLincoln Continental -1.023e+01 1.279e+00 -7.998 4.34e-08 ***\nmodelLotus Europa 2.341e+01 3.392e-01 69.015 < 2e-16 ***\nmodelMaserati Bora -1.408e+01 1.903e-01 -74.006 < 2e-16 ***\nmodelMazda RX4 -8.359e+00 1.638e-01 -51.017 < 2e-16 ***\nmodelMazda RX4 Wag -1.030e+01 1.761e-01 -58.494 < 2e-16 ***\nmodelMerc 230 2.481e+00 5.506e-02 45.064 < 2e-16 ***\nmodelMerc 240D 3.804e+00 5.586e-02 68.099 < 2e-16 ***\nmodelMerc 280 -2.984e+00 6.794e-02 -43.926 < 2e-16 ***\nmodelMerc 280C -4.382e+00 6.668e-02 -65.723 < 2e-16 ***\nmodelMerc 450SE -1.661e+00 1.075e-01 -15.448 1.26e-13 ***\nmodelMerc 450SL 7.892e-01 5.311e-02 14.861 2.83e-13 ***\nmodelMerc 450SLC -1.524e+00 6.416e-02 -23.749 < 2e-16 ***\nmodelPontiac Firebird 2.178e+00 7.002e-02 31.102 < 2e-16 ***\nmodelPorsche 914-2 8.306e+00 1.409e-01 58.945 < 2e-16 ***\nmodelToyota Corolla 1.419e+01 2.372e-01 59.809 < 2e-16 ***\nmodelToyota Corona 1.342e+01 2.208e-01 60.795 < 2e-16 ***\nmodelValiant 2.760e+00 1.050e-02 262.897 < 2e-16 ***\nmodelVolvo 142E -9.189e+00 1.720e-01 -53.428 < 2e-16 ***\namTRUE 1.302e+01 1.792e-01 72.629 < 2e-16 ***\ngear.L 1.571e-01 2.703e-02 5.811 6.42e-06 ***\ngear.Q -5.584e+00 4.818e-02 -115.914 < 2e-16 ***\ncarb -3.135e-04 4.119e-03 -0.076 0.939977 \n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nApproximate significance of smooth terms:\n edf Ref.df F p-value \ns(wt) 8.027 8.693 449 <2e-16 ***\n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nRank: 42/45\nR-sq.(adj) = 1 Deviance explained = 100%\nGCV = 0.00017704 Scale est. = 6.3549e-05 n = 64"
attr(,"params")$y_col
[1] "mpg"
attr(,"params")$rand_it
[1] 3
attr(,"params")$parallel
[1] 0
attr(,"params")$model_packages
NULL
attr(,"params")$random_model_call_string
NULL
attr(,"params")$random_model_call_string_vars
character(0)
attr(,"params")$positive
[1] TRUE
attr(,"params")$seed
[1] 0
attr(,"params")$rand_it_ok
[1] 100
attr(,"params")$exactness
[1] "surrogate"
Code
unclass(set_names(map(stats_names, function(.stat) {
p_to_random_value(pd@rand_stats$mpg, .stat, test_p)
}), stats_names))
Output
$aled
0 0.001 0.01 0.01 0.05 0.1
2.145276e-02 2.092763e-02 1.620142e-02 1.620142e-02 1.331799e-02 1.160576e-02
0.5 1
5.855495e-03 2.343212e-05
$aler_min
0 0.001 0.01 0.01 0.05
-8.018319e-02 -7.889883e-02 -6.733957e-02 -6.733957e-02 -4.856189e-02
0.1 0.5 1
-4.264222e-02 -1.985118e-02 -8.389436e-05
$aler_max
0 0.001 0.01 0.01 0.05 0.1
6.705638e-02 6.689973e-02 6.548988e-02 6.548988e-02 5.404877e-02 5.006009e-02
0.5 1
2.010266e-02 5.471472e-05
$naled
0 0.001 0.01 0.01 0.05 0.1 0.5 1
1.3916016 1.3916016 1.3916016 1.3916016 1.2207031 1.2207031 0.7080078 0.0000000
$naler_min
0 0.001 0.01 0.01 0.05 0.1 0.5 1
-1.5625 -1.5625 -1.5625 -1.5625 -1.5625 -1.5625 -1.5625 0.0000
$naler_max
0 0.001 0.01 0.01 0.05 0.1 0.5 1
3.1250 3.1250 3.1250 3.1250 3.1250 3.1250 1.5625 0.0000
ALEpDist works with custom random_model_call_string
Code
unclass(pd)
Output
<object>
attr(,"S7_class")
<ale::ALEpDist> class
@ parent : <S7_object>
@ constructor: function(model, data, ..., y_col, rand_it, surrogate, parallel, model_packages, random_model_call_string, random_model_call_string_vars, positive, pred_fun, pred_type, output_residuals, seed, silent, .skip_validation) {...}
@ validator : <NULL>
@ properties :
$ rand_stats : <list>
$ residual_distribution: S3<univariateML>
$ residuals : <double> or <NULL>
$ params : <list>
attr(,"rand_stats")
attr(,"rand_stats")$mpg
# A tibble: 3 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 0.000484 -0.00330 0.00201 0 0 0
2 0.00211 -0.00659 0.00835 0.342 -1.56 1.56
3 0.00196 -0.00644 0.00866 0.220 -1.56 1.56
attr(,"residual_distribution")
Maximum likelihood estimates for the Laplace model
mu sigma
1.303e-11 3.587e-03
attr(,"residuals")
[1] -9.470698e-04 -1.130145e-03 -3.078035e-03 7.415332e-04 -4.678952e-03
[6] 7.516518e-04 2.728091e-03 -8.853029e-03 -3.016706e-04 -1.794893e-03
[11] -3.673897e-03 -2.816578e-03 5.414042e-03 2.979146e-03 2.219206e-03
[16] 8.324011e-04 -4.976941e-05 -1.115543e-02 1.437735e-03 2.200997e-03
[21] 2.625747e-03 5.178720e-04 -9.802341e-03 7.118944e-03 5.255702e-03
[26] -9.746617e-03 -2.976337e-03 6.542735e-03 -8.071930e-03 4.016990e-03
[31] -2.747836e-04 -6.032700e-05 9.470698e-04 1.130145e-03 3.078035e-03
[36] -7.415332e-04 4.678952e-03 -7.516518e-04 -2.728091e-03 8.853029e-03
[41] 3.016706e-04 1.794893e-03 3.673897e-03 2.816578e-03 -5.414042e-03
[46] -2.979146e-03 -2.219206e-03 -8.324011e-04 4.976944e-05 1.115543e-02
[51] -1.437735e-03 -2.200997e-03 -2.625747e-03 -5.178720e-04 9.802341e-03
[56] -7.118944e-03 -5.255702e-03 9.746617e-03 2.976337e-03 -6.542735e-03
[61] 8.071930e-03 -4.016990e-03 2.747836e-04 6.032702e-05
attr(,"params")
attr(,"params")$model
attr(,"params")$model$class
[1] "gam" "glm" "lm"
attr(,"params")$model$call
[1] "mgcv::gam(formula = mpg ~ model + s(wt) + am + gear + carb, data = test_cars)"
attr(,"params")$model$print
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nEstimated degrees of freedom:\n8.03 total = 41.03 \n\nGCV score: 0.0001770391 rank: 42/45"
attr(,"params")$model$summary
[1] "\nFamily: gaussian \nLink function: identity \n\nFormula:\nmpg ~ model + s(wt) + am + gear + carb\n\nParametric coefficients:\n Estimate Std. Error t value Pr(>|t|) \n(Intercept) 1.432e+01 1.353e-01 105.784 < 2e-16 ***\nmodelCadillac Fleetwood -9.910e+00 1.259e+00 -7.873 5.68e-08 ***\nmodelCamaro Z28 -3.700e+00 7.268e-02 -50.911 < 2e-16 ***\nmodelChrysler Imperial -5.777e+00 1.276e+00 -4.526 0.000152 ***\nmodelDatsun 710 -3.793e+00 1.131e-01 -33.550 < 2e-16 ***\nmodelDodge Challenger -1.266e-01 2.060e-02 -6.147 2.87e-06 ***\nmodelDuster 360 -1.547e+00 2.851e-02 -54.276 < 2e-16 ***\nmodelFerrari Dino -4.088e+00 1.542e-01 -26.506 < 2e-16 ***\nmodelFiat 128 7.211e+00 9.518e-02 75.763 < 2e-16 ***\nmodelFiat X1-9 5.916e+00 1.941e-01 30.488 < 2e-16 ***\nmodelFord Pantera L -1.094e+01 1.737e-01 -63.000 < 2e-16 ***\nmodelHonda Civic 1.474e+01 2.896e-01 50.893 < 2e-16 ***\nmodelHornet 4 Drive 7.569e+00 5.315e-02 142.406 < 2e-16 ***\nmodelHornet Sportabout 3.468e+00 9.616e-03 360.698 < 2e-16 ***\nmodelLincoln Continental -1.023e+01 1.279e+00 -7.998 4.34e-08 ***\nmodelLotus Europa 2.341e+01 3.392e-01 69.015 < 2e-16 ***\nmodelMaserati Bora -1.408e+01 1.903e-01 -74.006 < 2e-16 ***\nmodelMazda RX4 -8.359e+00 1.638e-01 -51.017 < 2e-16 ***\nmodelMazda RX4 Wag -1.030e+01 1.761e-01 -58.494 < 2e-16 ***\nmodelMerc 230 2.481e+00 5.506e-02 45.064 < 2e-16 ***\nmodelMerc 240D 3.804e+00 5.586e-02 68.099 < 2e-16 ***\nmodelMerc 280 -2.984e+00 6.794e-02 -43.926 < 2e-16 ***\nmodelMerc 280C -4.382e+00 6.668e-02 -65.723 < 2e-16 ***\nmodelMerc 450SE -1.661e+00 1.075e-01 -15.448 1.26e-13 ***\nmodelMerc 450SL 7.892e-01 5.311e-02 14.861 2.83e-13 ***\nmodelMerc 450SLC -1.524e+00 6.416e-02 -23.749 < 2e-16 ***\nmodelPontiac Firebird 2.178e+00 7.002e-02 31.102 < 2e-16 ***\nmodelPorsche 914-2 8.306e+00 1.409e-01 58.945 < 2e-16 ***\nmodelToyota Corolla 1.419e+01 2.372e-01 59.809 < 2e-16 ***\nmodelToyota Corona 1.342e+01 2.208e-01 60.795 < 2e-16 ***\nmodelValiant 2.760e+00 1.050e-02 262.897 < 2e-16 ***\nmodelVolvo 142E -9.189e+00 1.720e-01 -53.428 < 2e-16 ***\namTRUE 1.302e+01 1.792e-01 72.629 < 2e-16 ***\ngear.L 1.571e-01 2.703e-02 5.811 6.42e-06 ***\ngear.Q -5.584e+00 4.818e-02 -115.914 < 2e-16 ***\ncarb -3.135e-04 4.119e-03 -0.076 0.939977 \n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nApproximate significance of smooth terms:\n edf Ref.df F p-value \ns(wt) 8.027 8.693 449 <2e-16 ***\n---\nSignif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nRank: 42/45\nR-sq.(adj) = 1 Deviance explained = 100%\nGCV = 0.00017704 Scale est. = 6.3549e-05 n = 64"
attr(,"params")$y_col
[1] "mpg"
attr(,"params")$rand_it
[1] 3
attr(,"params")$parallel
[1] 0
attr(,"params")$model_packages
NULL
attr(,"params")$random_model_call_string
[1] "mgcv::gam(\n mpg ~ model + s(wt) + am + gear + carb + random_variable,\n data = it.rand_data\n )"
attr(,"params")$random_model_call_string_vars
[1] "rmcsv"
attr(,"params")$positive
[1] TRUE
attr(,"params")$seed
[1] 0
attr(,"params")$rand_it_ok
[1] 3
attr(,"params")$exactness
[1] "invalid"
ALEpDist works with binary outcome
Code
unclass(pd)
Output
<object>
attr(,"S7_class")
<ale::ALEpDist> class
@ parent : <S7_object>
@ constructor: function(model, data, ..., y_col, rand_it, surrogate, parallel, model_packages, random_model_call_string, random_model_call_string_vars, positive, pred_fun, pred_type, output_residuals, seed, silent, .skip_validation) {...}
@ validator : <NULL>
@ properties :
$ rand_stats : <list>
$ residual_distribution: S3<univariateML>
$ residuals : <double> or <NULL>
$ params : <list>
attr(,"rand_stats")
attr(,"rand_stats")$vs
# A tibble: 10 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 5.16e-25 -1.32e-24 9.89e-25 3.64 -7.81 6.25
2 2.70e-25 -6.53e-25 6.73e-25 2.56 -1.56 4.69
3 1.43e-25 -3.61e-25 3.58e-25 2.22 -1.56 4.69
4 2.86e-25 -5.04e-25 4.75e-25 2.81 -1.56 4.69
5 8.00e-23 -1.39e-22 1.57e-22 23.8 -50 6.25
6 8.44e-26 -2.01e-25 1.96e-25 2.17 -1.56 3.12
7 1.12e-23 -2.59e-23 2.06e-23 22.8 -50 6.25
8 1.28e-21 -2.04e-21 2.84e-21 26.1 -50 6.25
9 0 0 0 0 0 0
10 6.51e-24 -1.16e-23 1.15e-23 19.7 -50 6.25
attr(,"residual_distribution")
Maximum likelihood estimates for the Uniform model
min max
-3.926e-13 3.926e-13
attr(,"params")
attr(,"params")$model
attr(,"params")$model$class
[1] "gam" "glm" "lm"
attr(,"params")$model$call
[1] "mgcv::gam(formula = vs ~ model + s(wt) + am + gear + carb, family = stats::binomial(), \n data = test_cars)"
attr(,"params")$model$print
[1] "\nFamily: binomial \nLink function: logit \n\nFormula:\nvs ~ model + s(wt) + am + gear + carb\n\nEstimated degrees of freedom:\n1 total = 34 \n\nUBRE score: 0.0625 rank: 42/45"
attr(,"params")$model$summary
[1] "\nFamily: binomial \nLink function: logit \n\nFormula:\nvs ~ model + s(wt) + am + gear + carb\n\nParametric coefficients:\n Estimate Std. Error z value Pr(>|z|)\n(Intercept) 9.522e+00 1.068e+06 0 1\nmodelCadillac Fleetwood 1.310e-09 2.420e+07 0 1\nmodelCamaro Z28 2.970e-10 5.648e+06 0 1\nmodelChrysler Imperial 1.380e-09 2.550e+07 0 1\nmodelDatsun 710 7.775e-08 9.376e+06 0 1\nmodelDodge Challenger 7.967e-11 1.760e+06 0 1\nmodelDuster 360 1.078e-10 2.390e+06 0 1\nmodelFerrari Dino -5.713e+01 1.689e+07 0 1\nmodelFiat 128 4.992e-08 7.891e+06 0 1\nmodelFiat X1-9 8.312e-08 4.293e+06 0 1\nmodelFord Pantera L -5.713e+01 2.207e+07 0 1\nmodelHonda Civic 0.000e+00 0.000e+00 NaN NaN\nmodelHornet 4 Drive 5.713e+01 2.914e+06 0 1\nmodelHornet Sportabout 1.906e-11 1.054e+06 0 1\nmodelLincoln Continental 1.451e-09 2.680e+07 0 1\nmodelLotus Europa 0.000e+00 0.000e+00 NaN NaN\nmodelMaserati Bora -5.713e+01 2.726e+07 0 1\nmodelMazda RX4 -5.713e+01 1.339e+07 0 1\nmodelMazda RX4 Wag -5.713e+01 1.657e+07 0 1\nmodelMerc 230 -1.764e-05 1.169e+06 0 1\nmodelMerc 240D 0.000e+00 0.000e+00 NaN NaN\nmodelMerc 280 3.537e-08 3.387e+06 0 1\nmodelMerc 280C -1.763e-05 3.113e+06 0 1\nmodelMerc 450SE 4.560e-10 8.472e+06 0 1\nmodelMerc 450SL 2.063e-10 3.993e+06 0 1\nmodelMerc 450SLC 2.581e-10 4.887e+06 0 1\nmodelPontiac Firebird 2.934e-10 5.495e+06 0 1\nmodelPorsche 914-2 -5.713e+01 8.485e+06 0 1\nmodelToyota Corolla -1.764e-05 3.047e+06 0 1\nmodelToyota Corona 5.713e+01 1.270e+07 0 1\nmodelValiant 5.713e+01 1.128e+06 0 1\nmodelVolvo 142E 8.495e-07 1.543e+07 0 1\namTRUE -6.645e-09 2.111e+07 0 1\ngear.L 4.040e+01 3.334e+06 0 1\ngear.Q -2.332e+01 7.942e+05 0 1\ncarb 3.703e-12 4.368e+05 0 1\n\nApproximate significance of smooth terms:\n edf Ref.df Chi.sq p-value\ns(wt) 1 1 0 1\n\nRank: 42/45\nR-sq.(adj) = 1 Deviance explained = 100%\nUBRE = 0.0625 Scale est. = 1 n = 64"
attr(,"params")$y_col
[1] "vs"
attr(,"params")$rand_it
[1] 10
attr(,"params")$parallel
[1] 0
attr(,"params")$model_packages
NULL
attr(,"params")$random_model_call_string
NULL
attr(,"params")$random_model_call_string_vars
character(0)
attr(,"params")$positive
[1] TRUE
attr(,"params")$seed
[1] 0
attr(,"params")$rand_it_ok
[1] 10
attr(,"params")$exactness
[1] "invalid"
ALEpDist works with categorical outcome
Code
unclass(pd)
Output
<object>
attr(,"S7_class")
<ale::ALEpDist> class
@ parent : <S7_object>
@ constructor: function(model, data, ..., y_col, rand_it, surrogate, parallel, model_packages, random_model_call_string, random_model_call_string_vars, positive, pred_fun, pred_type, output_residuals, seed, silent, .skip_validation) {...}
@ validator : <NULL>
@ properties :
$ rand_stats : <list>
$ residual_distribution: S3<univariateML>
$ residuals : <double> or <NULL>
$ params : <list>
attr(,"rand_stats")
attr(,"rand_stats")$Asia
# A tibble: 10 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 0 0 0 0 0 0
2 0 0 0 0 0 0
3 0 0 0 0 0 0
4 0 0 0 0 0 0
5 0 0 0 0 0 0
6 0 0 0 0 0 0
7 0 0 0 0 0 0
8 0 0 0 0 0 0
9 0 0 0 0 0 0
10 0 0 0 0 0 0
attr(,"rand_stats")$Europe
# A tibble: 10 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 0 0 0 0 0 0
2 0 0 0 0 0 0
3 0 0 0 0 0 0
4 0 0 0 0 0 0
5 0 0 0 0 0 0
6 0 0 0 0 0 0
7 0 0 0 0 0 0
8 0 0 0 0 0 0
9 0 0 0 0 0 0
10 0 0 0 0 0 0
attr(,"rand_stats")$`North America`
# A tibble: 10 x 6
aled aler_min aler_max naled naler_min naler_max
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 0 0 0 0 0 0
2 0 0 0 0 0 0
3 0 0 0 0 0 0
4 0 0 0 0 0 0
5 0 0 0 0 0 0
6 0 0 0 0 0 0
7 0 0 0 0 0 0
8 0 0 0 0 0 0
9 0 0 0 0 0 0
10 0 0 0 0 0 0
attr(,"residual_distribution")
Maximum likelihood estimates for the Laplace model
mu sigma
-2.043e-23 1.503e-17
attr(,"params")
attr(,"params")$model
attr(,"params")$model$class
[1] "multinom" "nnet"
attr(,"params")$model$call
[1] "nnet::multinom(formula = continent ~ model + wt + am + gear + \n carb, data = test_cars, trace = FALSE)"
attr(,"params")$model$print
[1] "Call:\nnnet::multinom(formula = continent ~ model + wt + am + gear + \n carb, data = test_cars, trace = FALSE)\n\nCoefficients:\n (Intercept) modelCadillac Fleetwood modelCamaro Z28\nEurope -21.86205 -2.222771 -6.526356\nNorth America -12.29302 -2.006450 14.817906\n modelChrysler Imperial modelDatsun 710 modelDodge Challenger\nEurope -3.5495185 -42.28844 -3.766810\nNorth America -0.4119232 -22.35666 7.689393\n modelDuster 360 modelFerrari Dino modelFiat 128 modelFiat X1-9\nEurope -7.046867 5.545836 42.144872 43.426334\nNorth America 20.851301 -3.241914 2.522922 2.870462\n modelFord Pantera L modelHonda Civic modelHornet 4 Drive\nEurope -23.11127 -27.5878163 -2.495884\nNorth America 29.30682 -0.1513286 7.771946\n modelHornet Sportabout modelLincoln Continental modelLotus Europa\nEurope -5.036042 -1.780289 14.561696\nNorth America 12.103127 -2.891703 -1.631007\n modelMaserati Bora modelMazda RX4 modelMazda RX4 Wag\nEurope 6.999218 -33.61112 -35.88135\nNorth America -7.857123 -3.27005 -10.72975\n modelMerc 230 modelMerc 240D modelMerc 280 modelMerc 280C\nEurope 14.2684900 14.2452958 11.590028 10.4496565\nNorth America -0.5204999 -0.6708756 1.228568 0.8928993\n modelMerc 450SE modelMerc 450SL modelMerc 450SLC\nEurope 39.49499 34.5616 36.09627\nNorth America -38.59611 -31.8537 -34.17067\n modelPontiac Firebird modelPorsche 914-2 modelToyota Corolla\nEurope -3.057543 15.344405 -34.531039\nNorth America 4.399831 -4.102117 -8.804816\n modelToyota Corona modelValiant modelVolvo 142E wt\nEurope -44.72605 -2.073514 34.932615 4.534617\nNorth America -52.77315 5.537827 2.626442 3.625818\n amTRUE gear.L gear.Q carb\nEurope -34.05607 40.7995829 -5.443825 10.524876\nNorth America -24.81811 0.6220113 39.516400 5.894396\n\nResidual Deviance: 0.0001161393 \nAIC: 136.0001 "
attr(,"params")$model$summary
[1] "Call:\nnnet::multinom(formula = continent ~ model + wt + am + gear + \n carb, data = test_cars, trace = FALSE)\n\nCoefficients:\n (Intercept) modelCadillac Fleetwood modelCamaro Z28\nEurope -21.86205 -2.222771 -6.526356\nNorth America -12.29302 -2.006450 14.817906\n modelChrysler Imperial modelDatsun 710 modelDodge Challenger\nEurope -3.5495185 -42.28844 -3.766810\nNorth America -0.4119232 -22.35666 7.689393\n modelDuster 360 modelFerrari Dino modelFiat 128 modelFiat X1-9\nEurope -7.046867 5.545836 42.144872 43.426334\nNorth America 20.851301 -3.241914 2.522922 2.870462\n modelFord Pantera L modelHonda Civic modelHornet 4 Drive\nEurope -23.11127 -27.5878163 -2.495884\nNorth America 29.30682 -0.1513286 7.771946\n modelHornet Sportabout modelLincoln Continental modelLotus Europa\nEurope -5.036042 -1.780289 14.561696\nNorth America 12.103127 -2.891703 -1.631007\n modelMaserati Bora modelMazda RX4 modelMazda RX4 Wag\nEurope 6.999218 -33.61112 -35.88135\nNorth America -7.857123 -3.27005 -10.72975\n modelMerc 230 modelMerc 240D modelMerc 280 modelMerc 280C\nEurope 14.2684900 14.2452958 11.590028 10.4496565\nNorth America -0.5204999 -0.6708756 1.228568 0.8928993\n modelMerc 450SE modelMerc 450SL modelMerc 450SLC\nEurope 39.49499 34.5616 36.09627\nNorth America -38.59611 -31.8537 -34.17067\n modelPontiac Firebird modelPorsche 914-2 modelToyota Corolla\nEurope -3.057543 15.344405 -34.531039\nNorth America 4.399831 -4.102117 -8.804816\n modelToyota Corona modelValiant modelVolvo 142E wt\nEurope -44.72605 -2.073514 34.932615 4.534617\nNorth America -52.77315 5.537827 2.626442 3.625818\n amTRUE gear.L gear.Q carb\nEurope -34.05607 40.7995829 -5.443825 10.524876\nNorth America -24.81811 0.6220113 39.516400 5.894396\n\nStd. Errors:\n (Intercept) modelCadillac Fleetwood modelCamaro Z28\nEurope 35852.83 64.65729 NaN\nNorth America 12968.78 64.70730 2.870067e-08\n modelChrysler Imperial modelDatsun 710 modelDodge Challenger\nEurope 477.3874 1.590757e-08 2.858243e-08\nNorth America 477.3874 2.200012e-18 6.428719e+00\n modelDuster 360 modelFerrari Dino modelFiat 128 modelFiat X1-9\nEurope 5.486580e-10 2660.898 1.805399e+04 1.525317e+04\nNorth America 7.249665e-11 2660.898 1.192455e-12 4.867466e-13\n modelFord Pantera L modelHonda Civic modelHornet 4 Drive\nEurope 0.09603878 5.871595e-07 3.863891e-10\nNorth America 0.09624948 9.331882e-08 7.163294e+03\n modelHornet Sportabout modelLincoln Continental modelLotus Europa\nEurope 4.655689e-09 285.0759 32194.07\nNorth America 5.351209e-02 285.1436 20099.45\n modelMaserati Bora modelMazda RX4 modelMazda RX4 Wag\nEurope 0.01739745 644.72277405 1.000913e+02\nNorth America 0.01739745 0.04282998 3.394594e-05\n modelMerc 230 modelMerc 240D modelMerc 280 modelMerc 280C\nEurope 1.329222e+02 1.088958e+02 4.023889e-07 2.701645e-06\nNorth America 4.475756e-08 3.800080e-08 2.855911e-10 1.280090e-09\n modelMerc 450SE modelMerc 450SL modelMerc 450SLC\nEurope 92448.24826 22918.50 1179.3388\nNorth America 12.05893 21677.09 776.0876\n modelPontiac Firebird modelPorsche 914-2 modelToyota Corolla\nEurope 1.504060e-06 29024.38 9.393802e-15\nNorth America 6.422677e+01 62881.87 2.008891e-13\n modelToyota Corona modelValiant modelVolvo 142E wt amTRUE\nEurope 2.042457e-17 7.332842e-09 5.751337e+04 26852.73 81773.79\nNorth America 1.689525e+01 3.329289e+04 5.749185e-12 97682.05 57698.40\n gear.L gear.Q carb\nEurope 120502.95 45794.583 77388.60\nNorth America 84695.32 5294.488 34719.66\n\nResidual Deviance: 0.0001161393 \nAIC: 136.0001 "
attr(,"params")$y_col
[1] "continent"
attr(,"params")$rand_it
[1] 10
attr(,"params")$parallel
[1] 0
attr(,"params")$model_packages
[1] "nnet"
attr(,"params")$random_model_call_string
NULL
attr(,"params")$random_model_call_string_vars
character(0)
attr(,"params")$positive
[1] TRUE
attr(,"params")$seed
[1] 0
attr(,"params")$rand_it_ok
[1] 10
attr(,"params")$exactness
[1] "invalid"
Try the ale package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.