Crate rooc

Expand description

§ROOC

ROOC is a modeling language for defining and solving optimization problems.

Write everything as a single source file:

use indexmap::IndexMap;
use rooc::pipe::{AutoSolverPipe, LinearModelPipe, ModelPipe, PipeContext, PipeRunner, PipeableData, PreModelPipe};
let source = "
max sum((value, i) in enumerate(values)) { value * x_i }
s.t.
   sum((weight, i) in enumerate(weights)) { weight * x_i } <= capacity
where
   let weights = [10, 60, 30, 40, 30, 20, 20, 2]
   let values = [1, 10, 15, 40, 60, 90, 100, 15]
   let capacity = 102
define
   x_i as Boolean for i in 0..len(weights)";
//use pipes to solve the problem
let pipe_runner = PipeRunner::new(vec![
   Box::new(rooc::pipe::CompilerPipe::new()),
   Box::new(PreModelPipe::new()),
   Box::new(ModelPipe::new()),
   Box::new(LinearModelPipe::new()),
   Box::new(AutoSolverPipe::new()),
]);
let result = pipe_runner
   .run(
       PipeableData::String(source.to_string()),
       &PipeContext::new(vec![], &IndexMap::new()),
   )
   .unwrap();
let last = result
   .into_iter()
   .last()
   .unwrap()
   .to_milp_solution()
       .unwrap();

println!("{}", last)

Or extend the language with your own functionality and data

Modules§

Macros§

bail_missing_token
bail_semantic_error
bail_wrong_argument
check_bounds
enum_with_variants_to_string
err_unexpected_token
make_pipe_error
match_or_bail
match_pipe_data_to
wrong_argument

Structs§

AddressableAccess: Represents array-like access to a variable, such as x[1][2].
Assignment: Represents a variable assignment in a solution.
BlockFunction: A function that operates on a fixed set of expressions. This represents functions like min, max, avg that take a set of arguments and compute a single result.
BlockScopedFunction: A function that operates over a set of values with iteration variables. This represents functions like sum, product, etc. that iterate over a set.
CompilationError: Represents a compilation error with location information and error details.
CompoundVariable: Represents a variable with subscript indices, such as x_1_2.
Constant: A constant value with a name and a value
EqualityConstraint: Represents a linear equality constraint in standard form: ax = b where a is a vector of coefficients and b is a constant.
FunctionCall: Represents a function call with its arguments, name and source span.
Graph
GraphEdge
GraphNode
InputSpan: Represents a span of text in the input source, tracking location information.
IterableSet: Represents an iterable set expression in the intermediate language.
LinearConstraint: Represents a linear constraint in the form: coefficients * variables comparison_operator rhs
LinearModel: Represents a complete linear programming model including variables, constraints, and objective function.
Linearizer: Manages the linearization process for expressions and constraints.
LpSolution: Represents a solution to a linear programming problem.
NormalizationContext: Context for tracking the normalization process of converting constraints to standard form.
OptimalTableau
OptimalTableauWithSteps
PreConstraint: Represents a constraint in a mathematical programming problem.
PreObjective: Represents an optimization objective in a mathematical programming problem.
RoocParser: A parser for the Rooc optimization modeling language.
RoocSolver
SimplexStep
Spanned: A wrapper type that associates a value with its location in source code.
StandardLinearModel: Represents a linear optimization model in standard form.
Tableau
Tuple: A tuple type that holds an ordered collection of Primitive values.

Enums§

BinOp
BlockFunctionKind
BlockScopedFunctionKind
CanonicalTransformError
Comparison
IntOrBoolValue: Represents a variable value that can be either boolean or integer.
IterableKind: Represents different types of iterable collections in the system.
LinearModelError: Errors that can occur when manipulating a LinearModel.
LinearizationError
MILPValue: Represents a variable value that can be either boolean or integer.
Operator
OperatorError: Represents errors that can occur during operator application.
OptimizationType
ParseError: Represents different types of parsing errors that can occur during compilation.
PreExp: Represents an expression in the intermediate language before final transformation.
PreVariableType: Represents the type of a variable before type checking and transformation
Primitive: Represents a primitive value in the system.
PrimitiveKind
RoocSolverError
SimplexError
SolverError: Represents errors that can occur during linear programming problem solving.
StepAction
UnOp
VariableType: Represents the final, resolved type of a variable after being compiled

Traits§

ApplyOp: Trait for types that can have operators applied to them.
RoocFunction: Trait defining the interface for Rooc functions.
Spreadable: Trait for types that can be spread into a sequence of primitives.

Functions§

auto_solver: Solves a any kind of linear programming problem by picking the right solver for the model.
default_rooc_function_to_latex: Converts a function call to LaTeX format.
default_rooc_function_to_string: Converts a function call to a string representation.
default_type_check: The default type check implementation, it performs type checking of function arguments against expected types.
default_wrong_number_of_arguments: Creates an error for wrong number of arguments.
default_wrong_type: Creates a type error for wrong argument types.
divide_matrix_row_by
find_invalid_variables: Finds variables in a domain that don’t satisfy a validation condition.
float_eq: Checks if two numbers are the same within 5 decimal digits
float_ne: Checks if two numbers are different within 5 decimal digits
format_var: Formats a variable term with its coefficient for display.
make_constraints_map_from_assignment
normalize_constraint: Converts a linear constraint to standard form by adding slack or surplus variables as needed.
process_variables: Processes coefficients and variables to create Times expressions, filtering based on index.
process_variables_binary: Similar to process_variables but works with binary variables.
solve_binary_lp_problem: Solves a binary linear programming problem.
solve_integer_binary_lp_problem: Solves a mixed integer-binary linear programming problem.
solve_milp_lp_problem: Solves a mixed-integer linear programming problem using the MicroLP solver.
solve_real_lp_problem_clarabel: Solves a linear programming problem with real variables using the Clarabel solver.
solve_real_lp_problem_micro_lp: Solves a linear programming problem with real variables using the microlp solver.
solve_real_lp_problem_slow_simplex: Solves a linear programming problem with real variables using a basic simplex algorithm.
to_standard_form: Converts a linear programming model into standard form.

Type Aliases§

FunctionContextMap

Crate roocCopy item path

§ROOC

Modules§

Macros§

Structs§

Enums§

Traits§

Functions§

Type Aliases§

Crate rooc