Performance of (population) historical forecast
Usage
add_historical_pred(
test,
train,
max_score,
discrete = TRUE,
add_uniform = TRUE,
include_samples = FALSE,
n_samples = NULL
)Arguments
- test
Testing dataframe. The only requirements is that it contains a column "Score".
- train
Training dataframe. The only requirements is that it contains a column "Score".
- max_score
Maximum value that the score can take
- discrete
Whether to estimate a discrete or continuous historical forecast
- add_uniform
Whether to include samples from uniform distribution when computing a discrete historical forecast. This ensures that all states are visited.
- include_samples
Whether to return samples from the historical forecast in the output
- n_samples
If
include_samples=TRUE, how many samples to return. When NULL, the function return the training set.
Value
Dataframe test appended by the columns "lpd", "RPS" (or CRPS if discrete=FALSE) and optionally "Samples"
Details
The continuous historical forecast is calculated by considering the training set as samples from the predictive distribution.
Examples
max_score <- 100
train <- data.frame(Score = rbinom(1e2, max_score, 0.2))
test <- data.frame(Score = rbinom(1e2, max_score, 0.5))
add_historical_pred(test, train, max_score)
#> Score lpd RPS
#> 1 43 -5.303305 0.10812257
#> 2 60 -5.303305 0.19523700
#> 3 44 -5.303305 0.11245093
#> 4 49 -5.303305 0.13558526
#> 5 56 -5.303305 0.17215242
#> 6 41 -5.303305 0.09976436
#> 7 53 -5.303305 0.15588377
#> 8 45 -5.303305 0.11687879
#> 9 47 -5.303305 0.12603302
#> 10 54 -5.303305 0.16120715
#> 11 48 -5.303305 0.13075939
#> 12 52 -5.303305 0.15065988
#> 13 50 -5.303305 0.14051063
#> 14 50 -5.303305 0.14051063
#> 15 50 -5.303305 0.14051063
#> 16 55 -5.303305 0.16663003
#> 17 57 -5.303305 0.17777431
#> 18 53 -5.303305 0.15588377
#> 19 45 -5.303305 0.11687879
#> 20 48 -5.303305 0.13075939
#> 21 58 -5.303305 0.18349571
#> 22 56 -5.303305 0.17215242
#> 23 46 -5.303305 0.12140615
#> 24 43 -5.303305 0.10812257
#> 25 57 -5.303305 0.17777431
#> 26 40 -5.303305 0.09573451
#> 27 53 -5.303305 0.15588377
#> 28 53 -5.303305 0.15588377
#> 29 49 -5.303305 0.13558526
#> 30 53 -5.303305 0.15588377
#> 31 45 -5.303305 0.11687879
#> 32 50 -5.303305 0.14051063
#> 33 56 -5.303305 0.17215242
#> 34 50 -5.303305 0.14051063
#> 35 46 -5.303305 0.12140615
#> 36 58 -5.303305 0.18349571
#> 37 50 -5.303305 0.14051063
#> 38 51 -5.303305 0.14553551
#> 39 57 -5.303305 0.17777431
#> 40 51 -5.303305 0.14553551
#> 41 45 -5.303305 0.11687879
#> 42 47 -5.303305 0.12603302
#> 43 43 -5.303305 0.10812257
#> 44 61 -5.303305 0.20125690
#> 45 53 -5.303305 0.15588377
#> 46 43 -5.303305 0.10812257
#> 47 50 -5.303305 0.14051063
#> 48 42 -5.303305 0.10389372
#> 49 45 -5.303305 0.11687879
#> 50 54 -5.303305 0.16120715
#> 51 47 -5.303305 0.12603302
#> 52 47 -5.303305 0.12603302
#> 53 53 -5.303305 0.15588377
#> 54 46 -5.303305 0.12140615
#> 55 45 -5.303305 0.11687879
#> 56 58 -5.303305 0.18349571
#> 57 55 -5.303305 0.16663003
#> 58 46 -5.303305 0.12140615
#> 59 48 -5.303305 0.13075939
#> 60 55 -5.303305 0.16663003
#> 61 50 -5.303305 0.14051063
#> 62 44 -5.303305 0.11245093
#> 63 50 -5.303305 0.14051063
#> 64 42 -5.303305 0.10389372
#> 65 48 -5.303305 0.13075939
#> 66 48 -5.303305 0.13075939
#> 67 42 -5.303305 0.10389372
#> 68 53 -5.303305 0.15588377
#> 69 57 -5.303305 0.17777431
#> 70 54 -5.303305 0.16120715
#> 71 45 -5.303305 0.11687879
#> 72 45 -5.303305 0.11687879
#> 73 46 -5.303305 0.12140615
#> 74 51 -5.303305 0.14553551
#> 75 47 -5.303305 0.12603302
#> 76 47 -5.303305 0.12603302
#> 77 47 -5.303305 0.12603302
#> 78 56 -5.303305 0.17215242
#> 79 51 -5.303305 0.14553551
#> 80 55 -5.303305 0.16663003
#> 81 45 -5.303305 0.11687879
#> 82 48 -5.303305 0.13075939
#> 83 61 -5.303305 0.20125690
#> 84 53 -5.303305 0.15588377
#> 85 53 -5.303305 0.15588377
#> 86 56 -5.303305 0.17215242
#> 87 45 -5.303305 0.11687879
#> 88 41 -5.303305 0.09976436
#> 89 50 -5.303305 0.14051063
#> 90 43 -5.303305 0.10812257
#> 91 47 -5.303305 0.12603302
#> 92 55 -5.303305 0.16663003
#> 93 53 -5.303305 0.15588377
#> 94 37 -5.303305 0.08424197
#> 95 47 -5.303305 0.12603302
#> 96 46 -5.303305 0.12140615
#> 97 49 -5.303305 0.13558526
#> 98 48 -5.303305 0.13075939
#> 99 50 -5.303305 0.14051063
#> 100 56 -5.303305 0.17215242