Solving MPS Files

HPRLP solves MPS (Mathematical Programming System) models through JuMP's file reader. This is the recommended path after removal of the old build_from_mps helper.

Quick Start

using JuMP, HPRLP

# Step 1: Read model from MPS file
model = read_from_file("path/to/problem.mps")

# Step 2: Attach HPRLP
set_optimizer(model, HPRLP.Optimizer)

# Step 3: Configure solver parameters
set_attribute(model, "stoptol", 1e-4)
set_attribute(model, "use_gpu", true)

# Step 4: Optimize
optimize!(model)

# Optional: access the HPRLP-specific result struct
result = unsafe_backend(model).results

Working with MPS Files

Basic Usage

using JuMP, HPRLP

# Read the model
model = read_from_file("model.mps")

# Attach HPRLP
set_optimizer(model, HPRLP.Optimizer)

# Set up solver attributes
set_attribute(model, "stoptol", 1e-4)
set_attribute(model, "use_gpu", true)

# Solve
optimize!(model)

result = unsafe_backend(model).results

if result.status == "OPTIMAL"
    println("Found optimal solution!")
    println("Objective: ", result.primal_obj)
    println("Solution vector: ", result.x)
else
    println("Solver stopped with status: ", result.status)
end

With Custom Parameters

using JuMP, HPRLP

# Read model
model = read_from_file("large_problem.mps")

# Attach HPRLP
set_optimizer(model, HPRLP.Optimizer)

# Configure attributes
set_attribute(model, "stoptol", 1e-9)      # Higher accuracy
set_attribute(model, "time_limit", 3600.0) # 1 hour time limit
set_attribute(model, "use_gpu", true)      # Enable GPU
set_attribute(model, "verbose", true)      # Show progress
set_attribute(model, "warm_up", true)      # Enable warmup for accurate timing

# Solve
optimize!(model)

result = unsafe_backend(model).results

Common MPS Sources

NETLIB

The classic NETLIB LP test set:

• Download from: http://www.netlib.org/lp/data/
• Contains problems of various sizes and characteristics
• Standard benchmark for LP solvers

MIPLIB

Mixed-Integer Programming Library (LP relaxations):

• Download from: https://miplib.zib.de/
• Includes continuous LP problems
• Challenging real-world instances

Hans Mittelmann's Benchmark

Collection of LP and optimization problems:

• http://plato.asu.edu/ftp/lptestset/
• Regularly updated
• Various problem classes

Performance Tips

1. GPU vs CPU:

   • Use GPU for large problems (> 10,000 variables/constraints)
   • Use CPU for small to medium problems or when GPU is unavailable

2. Tolerance:

   • Use 1e-6 or 1e-8 for high-accuracy requirements

3. Time Limits:

   • Set reasonable time limits for batch processing
   • Default is 3600 seconds (1 hour)

4. Scaling:

   • Keep scaling enabled for better numerical stability
   • Disable only if you have pre-scaled data

See Also

• Parameters - Complete guide to solver parameters and configuration
• Output & Results - Understanding and interpreting solver results