Installation

1. Install Julia

Install Julia language. Julia 1.9 or later is required for the current HOPE package setup. A short video tutorial on how to download and install Julia is provided here.

2. Install HOPE

After registration in General, install the latest release by package name:

import Pkg
Pkg.add("HOPE")

Until the initial registration is accepted, or to install the development branch directly, use the repository URL:

import Pkg
Pkg.add(url = "https://github.com/HOPE-Model-Project/HOPE.jl")

Both commands install and precompile the Julia package. They do not download model cases, start dashboards or MCP services, install Python, or require a commercial solver license.

3. Get model cases

Model cases are maintained in the separate HOPEModelCases repository. Clone them to a user-selected location:

git clone https://github.com/HOPE-Model-Project/HOPEModelCases /path/to/HOPEModelCases

Model-case downloads are explicit and separate from package installation. Set HOPE_MODELCASES_PATH to that location before running a file-based case:

  • Linux / macOS: export HOPE_MODELCASES_PATH=/path/to/HOPEModelCases
  • Windows (PowerShell): $env:HOPE_MODELCASES_PATH = "C:\path\to\HOPEModelCases"

4. Solvers and source development

A normal package installation includes the open-source solvers HiGHS, Cbc, GLPK, and Clp. No commercial license is required.

For development from a source checkout, activate and instantiate the repository:

import Pkg
Pkg.activate(".")
Pkg.instantiate()

Commercial solver packages such as Gurobi, SCIP, and CPLEX are not installed by Pkg.instantiate() by default. If needed, add them manually while the HOPE project is active:

import Pkg
Pkg.activate(".")
Pkg.add("Gurobi")   # or "SCIP" / "CPLEX"

When you do this from an active HOPE environment, the commercial solver package is added to the HOPE project environment, not just Julia's global default environment.

5. Minimal self-contained example

This one-bus DART SCUC example uses only package dependencies and creates no files:

using HOPE

data = DARTSystemData(
    generators = [
        DARTGenerator(
            name = "unit",
            bus = "bus",
            pmax_mw = 100.0,
            variable_cost_per_mwh = 25.0,
            commitment_required = false,
        ),
    ],
    network = DARTNetwork(bus_names = ["bus"]),
)
forecast = DARTForecast(
    interval_hours = 1.0,
    load_mw = reshape([40.0], 1, 1),
    availability = ones(1, 1),
)
result = solve_dart_scuc(data, forecast, default_dart_state(data))

result.generation_mw

The expected dispatch is 40 MW. Larger GTEP, PCM, holistic, and EREC examples are maintained separately in HOPEModelCases.