Release of Version 0.5.0

.github/workflows/ci-cd-develop.yml → .github/workflows/ci-cd.yml

@@ -46,7 +46,7 @@ jobs:
         shell: bash # required for the source command
         run: | # change to "./gradlew check --info" for more debugging output.
           source /opt/conda/bin/activate gams
-          ./gradlew check
+          ./gradlew check --info
 
       - name: Archive test reports
         if: always()
@@ -65,16 +65,20 @@ jobs:
   deploy-staging:
     runs-on: ubuntu-latest
     needs: check
-    if: ${{ github.ref == 'refs/heads/develop' }}
+    if: ${{ github.ref == 'refs/heads/main' || github.ref == 'refs/heads/develop' }}
     steps:
-      - name: Clone repo (shallow)
+      - name: Cloning repo
         uses: actions/checkout@v4
         with:
           fetch-depth: 0
-      - name: Push to dokku
-        uses: dokku/github-action@master
+
+      - name: Deploy via SSH and Docker Compose
+        uses: appleboy/ssh-action@v1.2.5
         with:
-          branch: develop
-          # only extracting the server address to a secret avoids leaking it in the logs
-          git_remote_url: ssh://dokku@${{ secrets.DOKKU_SERVER_ADDRESS }}/toolbox-backend-staging
-          ssh_private_key: ${{ secrets.DOKKU_DEPLOYMENT_KEY }}
+          host: ${{ secrets.DEPLOY_HOST }}
+          username: ${{ secrets.DEPLOY_USER }}
+          key: ${{ secrets.DEPLOY_SSH_KEY }}
+          command_timeout: 60m
+          script: |
+            echo "Deploying ${{ github.event.repository.name }} to '${{ github.ref_name }}' environment"
+            sh /root/ProvideQ/${{ github.event.repository.name }}/deploy.sh ${{ github.ref_name }}

.gitignore

@@ -52,7 +52,10 @@ __pycache__/
 # Software license key for GAMS, used for a docker container with GAMS included
 gamslice.txt
 
-# Listing files compiled from our GAMS scripts
+# Don't include GAMS Data eXchange (GDX)
+*.gdx
+
+# Listing files compiled from our GAMS scripts files, which are generated when running GAMS solvers
 *.lst
 *.op2
 solvers/gams/**/225a

CITATION.bib

@@ -0,0 +1,8 @@
+@inproceedings{eichhorn2025provideq,
+  title={Provideq: A quantum optimization toolbox},
+  author={Eichhorn, Domenik and Poser, Nick and Schweikart, Maximilian and Schaefer, Ina},
+  booktitle={2025 IEEE International Conference on Quantum Software (QSW)},
+  pages={206--214},
+  year={2025},
+  organization={IEEE}
+}

Dockerfile

@@ -32,6 +32,14 @@ RUN jlink \
 FROM debian:bookworm-slim AS runner
 WORKDIR /app
 
+# Install system build tools, needed to build solver dependencies
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    gcc \
+    g++ \
+    python3-dev \
+    && rm -rf /var/lib/apt/lists/*
+
 COPY scripts scripts
 
 # Install GAMS with conda python environment

README.md

@@ -1,4 +1,5 @@
-[![Build](https://img.shields.io/github/actions/workflow/status/ProvideQ/toolbox-server/deploy-main.yml?style=for-the-badge)](https://github.com/ProvideQ/toolbox-server/actions/workflows/deploy-main.yml)
+[![Build](https://img.shields.io/github/actions/workflow/status/ProvideQ/toolbox-server/ci-cd.yml?branch=main&style=for-the-badge
+)](https://github.com/ProvideQ/toolbox-server/actions/workflows/ci-cd.yml)
 [![Release](https://img.shields.io/github/v/release/ProvideQ/toolbox-server?style=for-the-badge)](https://github.com/ProvideQ/toolbox-server/releases/)
 
 # ProvideQ Toolbox Server
@@ -86,6 +87,6 @@ To use this, enable GitHub Actions and configure the following secrets in the Gi
   3. Set `DOKKU_DEPLOYMENT_KEY` to the private key generated in step i.
 
 ## License
-Copyright (c) 2022 - 2023 ProvideQ
+Copyright (c) 2022 - 2026 ProvideQ
 
 This project is available under the [MIT License](./LICENSE).

build.gradle

@@ -6,12 +6,16 @@ plugins {
 }
 
 group = 'edu.kit.provideq'
-version = '0.4.1'
-sourceCompatibility = '17'
+version = '0.5.0'
+
+java {
+    sourceCompatibility = JavaVersion.VERSION_17
+    targetCompatibility = JavaVersion.VERSION_17
+}
 
 repositories {
     mavenCentral()
-    maven { url 'https://jitpack.io' }
+    maven { url = 'https://jitpack.io' }
 }
 
 dependencies {
@@ -35,9 +39,6 @@ tasks.named('test') {
     useJUnitPlatform()
 }
 
-// set Java language level
-targetCompatibility = JavaVersion.VERSION_17
-
 tasks.withType(JavaCompile).configureEach {
     // treat "unchecked" warnings as errors
     options.compilerArgs << '-Xlint:unchecked'

config/checkstyle/checkstyle.xml

@@ -74,7 +74,7 @@
         </module>
         <module name="NeedBraces">
             <property name="tokens"
-                      value="LITERAL_DO, LITERAL_ELSE, LITERAL_FOR, LITERAL_IF, LITERAL_WHILE"/>
+                      value="LITERAL_DO, LITERAL_ELSE, LITERAL_FOR, LITERAL_WHILE"/>
         </module>
         <module name="LeftCurly">
             <property name="tokens"

demonstrators/qiskit/molecule-energy/molecule-energy.py

@@ -0,0 +1,113 @@
+from qiskit import QuantumCircuit
+from qiskit_algorithms import VQE
+from qiskit_algorithms.optimizers import L_BFGS_B, SLSQP
+from qiskit.circuit.library import TwoLocal
+from qiskit_nature.second_q.algorithms import GroundStateEigensolver
+from qiskit_nature.second_q.circuit.library import HartreeFock, UCCSD
+from qiskit_nature.second_q.drivers import PySCFDriver
+from qiskit_nature.second_q.mappers import JordanWignerMapper, ParityMapper
+from qiskit.primitives import Estimator
+import matplotlib.pyplot as plt
+from io import StringIO
+import numpy as np
+import sys
+
+arg_count =  len(sys.argv) - 1
+if arg_count != 1:
+    raise ValueError(f'This script expects exactly 1 argument: the molecule, but got {arg_count}: {sys.argv[1:]}')
+molecule = sys.argv[1]
+
+driver = PySCFDriver(atom=molecule)
+problem = driver.run()
+
+mapper = ParityMapper(num_particles=problem.num_particles)
+
+optimizer = L_BFGS_B()
+
+estimator = Estimator()
+
+ansatz = UCCSD(
+    problem.num_spatial_orbitals,
+    problem.num_particles,
+    mapper,
+    initial_state=HartreeFock(
+        problem.num_spatial_orbitals,
+        problem.num_particles,
+        mapper,
+    ),
+)
+
+vqe = VQE(estimator, ansatz, optimizer)
+vqe.initial_point = [0] * ansatz.num_parameters
+
+algorithm = GroundStateEigensolver(mapper, vqe)
+
+electronic_structure_result = algorithm.solve(problem)
+electronic_structure_result.formatting_precision = 6
+
+driver = PySCFDriver(atom='H .0 .0 .0; H .0 .0 0.74279')
+problem = driver.run()
+
+mapper = JordanWignerMapper()
+
+optimizer = SLSQP(maxiter=10000, ftol=1e-9)
+
+estimator = Estimator()
+
+var_forms = [['ry', 'rz'], 'ry']
+entanglements = ['full', 'linear']
+entanglement_blocks = ['cx', 'cz', ['cx', 'cz']]
+depths = list(range(1, 11))
+
+reference_circuit = QuantumCircuit(4)
+reference_circuit.x(0)
+reference_circuit.x(2)
+
+results = np.zeros((len(depths), len(entanglements), len(var_forms), len(entanglement_blocks)))
+
+for i, d in enumerate(depths):
+    for j, e in enumerate(entanglements):
+        for k, vf in enumerate(var_forms):
+            for l, eb in enumerate(entanglement_blocks):
+                variational_form = TwoLocal(4, rotation_blocks=vf, entanglement_blocks=eb, entanglement=e, reps=d)
+
+                ansatz = reference_circuit.compose(variational_form)
+
+                vqe = VQE(estimator, ansatz, optimizer)
+                vqe.initial_point = [0] * ansatz.num_parameters
+
+                algorithm = GroundStateEigensolver(mapper, vqe)
+
+                electronic_structure_result = algorithm.solve(problem)
+
+                results[i, j, k, l] = electronic_structure_result.total_energies[0]
+
+fig1, axs1 = plt.subplots(2, 3, sharey=True, sharex=True)
+fig2, axs2 = plt.subplots(2, 3, sharey=True, sharex=True)
+
+fig1.supxlabel('Depth')
+fig1.supylabel('Estimated ground state energy')
+fig2.supxlabel('Depth')
+fig2.supylabel('Estimated ground state energy')
+
+for j, e in enumerate(entanglements):
+    for l, eb in enumerate(entanglement_blocks):
+        axs1[j, l].plot(depths, results[:, j, 0, l])
+        axs1[j, l].set_title(f'{e}, ryrz, {eb}')
+        axs1[j, l].text(0.90, 0.75, f'Min: {np.min(results[:, j, 0, l]):.3f}')
+
+for j, e in enumerate(entanglements):
+    for l, eb in enumerate(entanglement_blocks):
+        axs2[j, l].plot(depths, results[:, j, 1, l])
+        axs2[j, l].set_title(f'{e}, ry, {eb}')
+        axs2[j, l].text(0.90, 0.75, f'Min: {np.min(results[:, j, 1, l]):.3f}')
+
+
+def print_fig(fig):
+    string_io = StringIO()
+    fig.savefig(string_io, format='svg')
+    print(string_io.getvalue())
+
+
+print_fig(fig1)
+print_fig(fig2)

demonstrators/qiskit/molecule-energy/requirements.txt

@@ -0,0 +1,14 @@
+# This file describes the python package requirements for all Qiskit demonstrator scripts
+# supported by ProvideQ
+
+# required for molecule energy solver
+qiskit==1.1.0
+qiskit-aer==0.14.2
+qiskit-algorithms==0.3.0
+qiskit-ibm-runtime==0.25.0
+qiskit-machine-learning==0.7.2
+qiskit-nature==0.7.2
+qiskit-nature-pyscf==0.4.0
+qiskit-qasm3-import==0.5.0
+qiskit-transpiler-service==0.4.5
+matplotlib==3.9.2

scripts/install-python-dependencies.py

@@ -1,9 +1,14 @@
 import os
 import subprocess
 import platform
+import sys
 
 script_dir = os.path.dirname(os.path.realpath(__file__))
 root_dir = os.path.dirname(script_dir)
+
+# Check if a specific folder argument was provided
+folder_filter = sys.argv[1] if len(sys.argv) > 1 else None
+
 base_dirs = [
     os.path.join(root_dir, 'demonstrators'),
     os.path.join(root_dir, 'solvers'),
@@ -19,13 +24,17 @@
             # Iterate over problem directory (knapsack, tsp, etc.)
             for solver_name in os.listdir(framework_dir):
                 solver_dir = os.path.join(framework_dir, solver_name)
+                # If folder filter is specified, only process matching folders
+                if folder_filter and folder_filter not in solver_dir:
+                    continue
+
                 req_file = os.path.join(solver_dir, 'requirements.txt')
                 if os.path.exists(req_file):
                     venv_name = f"{os.path.basename(root)}_{framework_name}_{solver_name}"
                     print(f"Setting up virtual environment '{venv_name}' for {solver_dir}...")
                     try:
                         venv_path = os.path.join('venv', venv_name)
-                        subprocess.run(['python3', '-m', 'venv', venv_path], check=True)
+                        subprocess.run(['python', '-m', 'venv', venv_path], check=True)
                         if platform.system() == 'Windows':
                             pip_executable = os.path.join(venv_path, 'Scripts', 'pip.exe')
                         else:

scripts/setup-gams.sh

@@ -19,7 +19,7 @@ apt-get install curl --yes
 # iisaa/gams-docker (https://github.com/iiasa/gams-docker/blob/master/Dockerfile, GPL-3.0 licensed)
 
 # Configure the GAMS version here
-GAMS_VERSION_RELEASE_MAJOR=49.3
+GAMS_VERSION_RELEASE_MAJOR=52.5
 GAMS_VERSION_HOTFIX=0
 
 # download the self-extracting archive to /opt/gams/gams.exe and run/extract it

solvers/_utility/parse_qubo_from_lp.py

@@ -0,0 +1,34 @@
+def parse_qubo_from_lp(model) -> tuple[list[str], dict[tuple[int, int], float], dict[int, float], float]:
+    vars = [x.VarName for x in model.getVars() if x.VType == "B"]
+
+    obj = model.getObjective()
+
+    quadratic_terms = {
+        (vars.index(term[1].VarName), vars.index(term[2].VarName)): term[0]
+        for term in obj.quadTerms()
+    }
+
+    linear_terms = {
+        vars.index(var.VarName): coeff
+        for coeff, var in obj.linTerms()
+        if var.VType == "B"
+    }
+
+    # Detect constant terms from non-binary variables
+    # A variable is considered a constant if:
+    # 1. Its LB equals its UB (fixed variable), OR
+    # 2. Its name can be parsed as an integer
+    constant_offset = obj.getLinExpr().getConstant()
+
+    # Extract constant contributions
+    constant_contributions = [
+        coeff * lb if lb == ub else (
+            coeff * int(var.VarName) if var.VarName.isdigit() else 0
+        )
+        for coeff, var in obj.getLinExpr().linTerms()
+        for lb, ub in [(var.getAttr('LB'), var.getAttr('UB'))]
+    ]
+
+    constant_offset += sum(constant_contributions)
+
+    return vars, quadratic_terms, linear_terms, constant_offset