Skip to contents

Define a variable for an lp_problem()

Source: R/variables.R
lp_variable.Rd

Add a variable to a problem by defining it's name, dimensions, type, and bounds.

Usage

lp_variable(
  .problem,
  definition,
  integer = FALSE,
  binary = FALSE,
  lower = -Inf,
  upper = +Inf
)

lp_var(
  .problem,
  definition,
  integer = FALSE,
  binary = FALSE,
  lower = -Inf,
  upper = +Inf
)

Arguments

.problem

An lp_problem() object.

definition

Expression.

  • If the variable is a scalar, simply type it's name.

    • lp_variable(x)

  • If the variable is a vector, type it's name and indices. Indices can also be named.

    • lp_variable( v[1:5] )

    • lp_variable( v[letters[1:5]] )

    • lp_variable( v[ind = letters[1:5]] )

    • ind <- letters[1:5]; lp_variable( v[ind] )

    The last two have the same result.

  • If the variable is a matrix or n-dimensional array, type it's name and the indices for every dimension (base::dimnames()), comma-separated.

    • lp_variable( mat[1:2, 1:3] )

    • lp_variable( arr[1:2, 1:3, 1:2] )

integer

Boolean, whether to treat variable as integer.

binary

Boolean, whether to treat variable as binary, {0, 1}.

lower

Numeric scalar or array. Lower bound for the variable.

upper

Numeric scalar or array. Upper bound for the variable.

Value

The .problem with an added variable in $variables. The fields of lp_variable objects are intended for internal use, modifying them is highly discouraged.

  • $name : String, name of the variable.

  • $lower and $upper : Bounds.

  • $type : String, one of "real", "integer" or "binary".

  • $integer and $binary : Booleans. If $binary is true, then $integer is also true.

The following fields are meant for internal use only.

  • $ind : Integer array. Indicates which indices correspond to this variable. Meant for internal use only.

  • $coef : Numeric matrix. The number of rows is the length of the variable, the number of columns is the total amount of variables in the problem. The values represent coefficients that are modified when adding variables or multiplying by a constant. Used for objective, constraints and aliases.

  • $add : Numeric column vector. The number of rows is the length of the variable. The values represent addends, modified when adding the variable and a constant. Used for objective, constraints and aliases.

  • $raw : Boolean indicating if the variable has been modified in any way (indexed, multiplied, summed, ...) or remains as defined.

Examples

# Diet problem
## You can only eat as much as you have of each food.

pantry <- c(Cookies = 8, Bread = 5, Milk = 1, Cereal = 3)
food <- names(pantry)

lp_problem() |>
    lp_variable(diet[food], lower = 0, upper = pantry)
#> <lpsugar Linear Problem>
#> 
#> -- $variables --
#> $diet
#> Real variable 'diet[food]'
#> 
#> 


# Graph problem
## active[i,j] = 1  if the edge from i to j is active.

nodes <- paste("node", 1:10)

lp_problem() |>
    lp_variable(active[nodes, nodes], binary = TRUE)
#> <lpsugar Linear Problem>
#> 
#> -- $variables --
#> $active
#> Binary variable 'active[nodes, nodes]'
#> 0 <= active <= 1
#> 
#>