More on Sparse Variable/Constraint Generation

[sbmack]

Again, my curious interest is in pre-processing sparse data objects outside of MIPModel if possible. I created a simple sparse supply chain dataframe as an example:

nums <- str_pad(1:20, 2, pad = "0")
plant <- paste0("pl", nums[1:3])
prod <- paste0('pr', nums[1:6])
ware <- paste0('wh', nums[1:2])

mat1 <- expand.grid(plant, prod, ware)
set.seed(7)
active <- sample(0:1, nrow(mat1), replace = TRUE, prob = c(0.3, 0.7))
mat2 <- cbind(mat1, active)
colnames(mat3) <- c('plant', 'prod', 'ware', 'active')
mat3 <- dplyr::filter(mat2, active !=0)
nrows <- nrow(mat3)
cost <- sample(30:300, nrows, replace = TRUE)
mat4 <- cbind(mat3, cost)

In this case 3 plants produce 6 common products that are delivered to 2 warehouses. However not all plant, product, warehouse combinations are feasible. mat4 is a sparse dataframe of feasible combinations with a cost vector attached.

For this example, I am thinking that an ideal MIPModel solution for generating those variables would simply be:

model <- MIPModel() %>%
  add_variable(x[mat4[, 1:3]])

Where x is the sparse index set read in directly. I read the help for add_variable:

add_variable(.model, .variable, ..., type = "continuous", lb = -Inf, ub = Inf)

add_variable_(
  .model,
  .variable,
  ...,
  type = "continuous",
  lb = -Inf,
  ub = Inf,
  .dots
)

But it is not clear to me how a .dots filter should be constructed with the current github version of MIPModel to generate a sparse variable set as in this simple sample case.

Comments? Ideas?

[dirkschumacher]

add_variable_ is deprecated.

You can use "parallel sequences" from the {listcomp} package:

library(stringr)
library(ompr)
nums <- str_pad(1:20, 2, pad = "0")
plant <- paste0("pl", nums[1:3])
prod <- paste0('pr', nums[1:6])
ware <- paste0('wh', nums[1:2])

mat1 <- expand.grid(plant, prod, ware)
set.seed(7)
active <- sample(0:1, nrow(mat1), replace = TRUE, prob = c(0.3, 0.7))
mat2 <- cbind(mat1, active)
mat3 <- dplyr::filter(mat2, active !=0)
colnames(mat3) <- c('plant', 'prod', 'ware', 'active')
nrows <- nrow(mat3)
cost <- sample(30:300, nrows, replace = TRUE)
mat4 <- cbind(mat3, cost)

MIPModel() |> 
  add_variable(x[i, j, k], list(i = mat4$plant, j = mat4$prod, k = mat4$ware)) |> 
  variable_keys()
#>  [1] "x[pl02,pr01,wh01]" "x[pl03,pr01,wh01]" "x[pl01,pr02,wh01]"
#>  [4] "x[pl02,pr02,wh01]" "x[pl01,pr03,wh01]" "x[pl03,pr03,wh01]"
#>  [7] "x[pl01,pr04,wh01]" "x[pl02,pr04,wh01]" "x[pl03,pr04,wh01]"
#> [10] "x[pl02,pr05,wh01]" "x[pl03,pr05,wh01]" "x[pl01,pr06,wh01]"
#> [13] "x[pl02,pr06,wh01]" "x[pl03,pr06,wh01]" "x[pl02,pr01,wh02]"
#> [16] "x[pl03,pr01,wh02]" "x[pl01,pr02,wh02]" "x[pl02,pr03,wh02]"
#> [19] "x[pl03,pr03,wh02]" "x[pl01,pr04,wh02]" "x[pl03,pr04,wh02]"
#> [22] "x[pl01,pr05,wh02]" "x[pl02,pr05,wh02]" "x[pl03,pr05,wh02]"
#> [25] "x[pl01,pr06,wh02]" "x[pl02,pr06,wh02]"
mat4
#>    plant prod ware active cost
#> 1   pl02 pr01 wh01      1  129
#> 2   pl03 pr01 wh01      1  195
#> 3   pl01 pr02 wh01      1  277
#> 4   pl02 pr02 wh01      1  200
#> 5   pl01 pr03 wh01      1  153
#> 6   pl03 pr03 wh01      1  278
#> 7   pl01 pr04 wh01      1   32
#> 8   pl02 pr04 wh01      1  210
#> 9   pl03 pr04 wh01      1  117
#> 10  pl02 pr05 wh01      1   70
#> 11  pl03 pr05 wh01      1  210
#> 12  pl01 pr06 wh01      1  293
#> 13  pl02 pr06 wh01      1   70
#> 14  pl03 pr06 wh01      1   60
#> 15  pl02 pr01 wh02      1  215
#> 16  pl03 pr01 wh02      1  103
#> 17  pl01 pr02 wh02      1  203
#> 18  pl02 pr03 wh02      1  269
#> 19  pl03 pr03 wh02      1  143
#> 20  pl01 pr04 wh02      1   61
#> 21  pl03 pr04 wh02      1  224
#> 22  pl01 pr05 wh02      1  262
#> 23  pl02 pr05 wh02      1   84
#> 24  pl03 pr05 wh02      1  101
#> 25  pl01 pr06 wh02      1  207
#> 26  pl02 pr06 wh02      1   47

^{Created on 2022-02-03 by the reprex package (v2.0.1)}

[sbmack]

Unglaublich! Fantastisch! Du bist bestimmt Schöpfergeist! Danke...

[dirkschumacher]

You can mark discussion posts as “answered” I think. (Moved it to Q&A)