prepare.idbrms_convolution.RdPrepare data for fitting using a convolution model
# S3 method for idbrms_convolution
prepare(
data,
location,
primary,
secondary,
initial_obs = 14,
max_convolution = 30,
...
)| data | A data frame containing at least two integer observations and a date variable. |
|---|---|
| location | Character string, variable to use as the spatial location. |
| primary | Character string, variable to use as the primary observation. |
| secondary | Character string, variable to use as the secondary observation. |
| initial_obs | Integer, number of observations to hold out from the likelihood. This is useful as initially the outcome will depend on primary data outside of the range of the training set and including this could bias the estimated delay distribution. Defaults to 14 days. |
| max_convolution | Integer defining the maximum index to use for the convolution. Defaults to 30 days. |
| ... | Additional arguments passed to model specific prepare functions |
Sam Abbott
# define some example data
library(data.table)
dt <- data.table(
region = "France", cases = seq(10, 500, by = 10),
date = seq(as.Date("2020-10-01"), by = "days", length.out = 50)
)
dt <- dt[, deaths := as.integer(shift(cases, 5) * 0.1)]
dt <- dt[is.na(deaths), deaths := 0]
dt <- prepare(
dt, model = "convolution", location = "region",
primary = "cases", secondary = "deaths",
)
dt[]
#> location date time index init_obs cstart cmax primary secondary
#> 1: France 2020-10-01 0 1 1 1 1 10 0
#> 2: France 2020-10-02 1 2 1 1 2 20 0
#> 3: France 2020-10-03 2 3 1 1 3 30 0
#> 4: France 2020-10-04 3 4 1 1 4 40 0
#> 5: France 2020-10-05 4 5 1 1 5 50 0
#> 6: France 2020-10-06 5 6 1 1 6 60 1
#> 7: France 2020-10-07 6 7 1 1 7 70 2
#> 8: France 2020-10-08 7 8 1 1 8 80 3
#> 9: France 2020-10-09 8 9 1 1 9 90 4
#> 10: France 2020-10-10 9 10 1 1 10 100 5
#> 11: France 2020-10-11 10 11 1 1 11 110 6
#> 12: France 2020-10-12 11 12 1 1 12 120 7
#> 13: France 2020-10-13 12 13 1 1 13 130 8
#> 14: France 2020-10-14 13 14 1 1 14 140 9
#> 15: France 2020-10-15 14 15 0 1 15 150 10
#> 16: France 2020-10-16 15 16 0 1 16 160 11
#> 17: France 2020-10-17 16 17 0 1 17 170 12
#> 18: France 2020-10-18 17 18 0 1 18 180 13
#> 19: France 2020-10-19 18 19 0 1 19 190 14
#> 20: France 2020-10-20 19 20 0 1 20 200 15
#> 21: France 2020-10-21 20 21 0 1 21 210 16
#> 22: France 2020-10-22 21 22 0 1 22 220 17
#> 23: France 2020-10-23 22 23 0 1 23 230 18
#> 24: France 2020-10-24 23 24 0 1 24 240 19
#> 25: France 2020-10-25 24 25 0 1 25 250 20
#> 26: France 2020-10-26 25 26 0 1 26 260 21
#> 27: France 2020-10-27 26 27 0 1 27 270 22
#> 28: France 2020-10-28 27 28 0 1 28 280 23
#> 29: France 2020-10-29 28 29 0 1 29 290 24
#> 30: France 2020-10-30 29 30 0 1 30 300 25
#> 31: France 2020-10-31 30 31 0 1 31 310 26
#> 32: France 2020-11-01 31 32 0 2 31 320 27
#> 33: France 2020-11-02 32 33 0 3 31 330 28
#> 34: France 2020-11-03 33 34 0 4 31 340 29
#> 35: France 2020-11-04 34 35 0 5 31 350 30
#> 36: France 2020-11-05 35 36 0 6 31 360 31
#> 37: France 2020-11-06 36 37 0 7 31 370 32
#> 38: France 2020-11-07 37 38 0 8 31 380 33
#> 39: France 2020-11-08 38 39 0 9 31 390 34
#> 40: France 2020-11-09 39 40 0 10 31 400 35
#> 41: France 2020-11-10 40 41 0 11 31 410 36
#> 42: France 2020-11-11 41 42 0 12 31 420 37
#> 43: France 2020-11-12 42 43 0 13 31 430 38
#> 44: France 2020-11-13 43 44 0 14 31 440 39
#> 45: France 2020-11-14 44 45 0 15 31 450 40
#> 46: France 2020-11-15 45 46 0 16 31 460 41
#> 47: France 2020-11-16 46 47 0 17 31 470 42
#> 48: France 2020-11-17 47 48 0 18 31 480 43
#> 49: France 2020-11-18 48 49 0 19 31 490 44
#> 50: France 2020-11-19 49 50 0 20 31 500 45
#> location date time index init_obs cstart cmax primary secondary