Multigeometry and expressive boundaries

.github/workflows/tests.yml

@@ -23,7 +23,7 @@ jobs:
           cache: "pip" # caching pip dependencies
       - name: Install dependencies
         run: |
-          sudo apt-get install -y -qq libboost-all-dev cmake
+          sudo apt-get update && sudo apt-get install -y -qq libboost-all-dev cmake
       - name: Install qlbm
         run: |
           pip install --upgrade pip 

README.md

@@ -66,8 +66,8 @@ There are also Docker container images in the `Docker` directory that can be use
 
 Currently, `qlbm` supports two algorithms:
  - The Quantum Transport Method (Collisionless QLBM) described in [Efficient and fail-safe quantum algorithm for the transport equation](https://doi.org/10.1016/j.jcp.2024.112816) ([arXiv:2211.14269](https://arxiv.org/abs/2211.14269)) by M.A. Schalkers and M. Möller.
- - The Space-Time QLBM/QLGA described in [On the importance of data encoding in quantum Boltzmann methods](https://link.springer.com/article/10.1007/s11128-023-04216-6) by M.A. Schalkers and M. Möller and expanded in [Fully Quantum Lattice Gas Automata Building Blocks for Computational Basis State Encodings](https://arxiv.org/abs/2506.12662).
- - The Linear-encoding Quantum Lattice Gas Automata (LQLGA) described in [On quantum extensions of hydrodynamic lattice gas automata](https://www.mdpi.com/2410-3896/4/2/48) by P. Love and [Fully Quantum Lattice Gas Automata Building Blocks for Computational Basis State Encodings](https://arxiv.org/abs/2506.12662).
+ - The Space-Time QLBM/QLGA described in [On the importance of data encoding in quantum Boltzmann methods](https://link.springer.com/article/10.1007/s11128-023-04216-6) by M.A. Schalkers and M. Möller and expanded in [Fully Quantum Lattice Gas Automata Building Blocks for Computational Basis State Encodings](https://doi.org/10.1016/j.jcp.2025.114595).
+ - The Linear-encoding Quantum Lattice Gas Automata (LQLGA) described in [On quantum extensions of hydrodynamic lattice gas automata](https://www.mdpi.com/2410-3896/4/2/48) by P. Love and [Fully Quantum Lattice Gas Automata Building Blocks for Computational Basis State Encodings](https://doi.org/10.1016/j.jcp.2025.114595).
 
 The `demos` directory contains several use cases for simulating and analyzing these algorithms. Each demo requires minimal setup once the virtual environment has been configured. Consult the `README.md` file in the `demos` directory for further details.
 

demos/simulation/ab_simulation.ipynb

@@ -19,13 +19,13 @@
     "\n",
     "from qlbm.components import (\n",
     "    CQLBM,\n",
+    "    ABDiscreteUniformInitialConditions,\n",
     "    ABGridMeasurement,\n",
-    "    ABInitialConditions,\n",
     "    EmptyPrimitive,\n",
     ")\n",
     "from qlbm.infra import QiskitRunner, SimulationConfig\n",
     "from qlbm.lattice import ABLattice\n",
-    "from qlbm.tools.utils import create_directory_and_parents\n"
+    "from qlbm.tools.utils import create_directory_and_parents"
    ]
   },
   {
@@ -57,7 +57,11 @@
    "outputs": [],
    "source": [
     "cfg = SimulationConfig(\n",
-    "    initial_conditions=ABInitialConditions(lattice),\n",
+    "    initial_conditions=ABDiscreteUniformInitialConditions(\n",
+    "        lattice,\n",
+    "        [1, 2, 5],  # E, N, and NE\n",
+    "        ([], list(range(5))),  # Span the entire y axis, but nothing on the x axis\n",
+    "    ),\n",
     "    algorithm=CQLBM(lattice),\n",
     "    postprocessing=EmptyPrimitive(lattice),\n",
     "    measurement=ABGridMeasurement(lattice),\n",
@@ -67,7 +71,7 @@
     "    statevector_sampling=True,\n",
     "    execution_backend=AerSimulator(method=\"statevector\"),\n",
     "    sampling_backend=AerSimulator(method=\"statevector\"),\n",
-    ")\n"
+    ")"
    ]
   },
   {
@@ -115,6 +119,14 @@
     "    statevector_snapshots=True,\n",
     ")"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "b61274d7",
+   "metadata": {},
+   "outputs": [],
+   "source": []
   }
  ],
  "metadata": {

demos/simulation/ab_simulation_parallel_bcs.ipynb

@@ -0,0 +1,184 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "348de144",
+   "metadata": {},
+   "source": [
+    "# Simulating the Amplitude-Based Collisionless QLBM"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "181af34e",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from qiskit_aer import AerSimulator\n",
+    "\n",
+    "from qlbm.components import (\n",
+    "    CQLBM,\n",
+    "    ABDiscreteUniformInitialConditions,\n",
+    "    ABGridMeasurement,\n",
+    "    EmptyPrimitive,\n",
+    ")\n",
+    "from qlbm.infra import QiskitRunner, SimulationConfig\n",
+    "from qlbm.lattice import ABLattice\n",
+    "from qlbm.tools.utils import create_directory_and_parents"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "05e4915d",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "lattice = ABLattice(\n",
+    "    {\n",
+    "        \"lattice\": {\"dim\": {\"x\": 16, \"y\": 16}, \"velocities\": \"d2q9\"},\n",
+    "        \"geometry\": [],\n",
+    "    }\n",
+    ")\n",
+    "\n",
+    "output_dir = \"qlbm-output/ab-pbc-d2q9-16x16-1-obstacle-qiskit\"\n",
+    "create_directory_and_parents(output_dir)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "5e274b50",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Geometry of marker |0>: two \"stacked\" \"wide\" rectangles\n",
+    "# Geometry of the marker |1>: one \"tall\" rectangle under the first two\n",
+    "# These geometries will act fully in parallel, and we will observe the average behavior\n",
+    "lattice.set_geometries(\n",
+    "    [\n",
+    "        [\n",
+    "            {\"shape\": \"cuboid\", \"x\": [6, 12], \"y\": [12, 14], \"boundary\": \"bounceback\"},\n",
+    "            {\"shape\": \"cuboid\", \"x\": [6, 12], \"y\": [8, 10], \"boundary\": \"bounceback\"},\n",
+    "            # {\"shape\": \"cuboid\", \"x\": [6, 8], \"y\": [0, 6], \"boundary\": \"bounceback\"},\n",
+    "        ],\n",
+    "        [{\"shape\": \"cuboid\", \"x\": [6, 8], \"y\": [0, 6], \"boundary\": \"bounceback\"}],\n",
+    "    ]\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "89709eb9",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "lattice.geometries"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "deb472a0",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from qlbm.components.ab.initial import ABParallelDiscreteUniformInitialConditions\n",
+    "\n",
+    "cfg = SimulationConfig(\n",
+    "    initial_conditions=ABParallelDiscreteUniformInitialConditions(\n",
+    "        lattice,\n",
+    "        [[1], [1]],\n",
+    "        [([0, 1], [0, 1, 2, 3]), ([0, 1], [0, 1, 2, 3])],\n",
+    "    ),\n",
+    "    algorithm=CQLBM(lattice),\n",
+    "    postprocessing=EmptyPrimitive(lattice),\n",
+    "    measurement=ABGridMeasurement(lattice),\n",
+    "    target_platform=\"QISKIT\",\n",
+    "    compiler_platform=\"QISKIT\",\n",
+    "    optimization_level=0,\n",
+    "    statevector_sampling=True,\n",
+    "    execution_backend=AerSimulator(method=\"statevector\"),\n",
+    "    sampling_backend=AerSimulator(method=\"statevector\"),\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "57240678",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "cfg.prepare_for_simulation()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "1da18eb8",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Number of shots to simulate for each timestep when running the circuit\n",
+    "NUM_SHOTS = 2**12\n",
+    "\n",
+    "# Number of timesteps to simulate\n",
+    "NUM_STEPS = 20"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "5c8c7d1e",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "runner = QiskitRunner(\n",
+    "    cfg,\n",
+    "    lattice,\n",
+    ")\n",
+    "\n",
+    "\n",
+    "# Simulate the circuits using both snapshots\n",
+    "runner.run(\n",
+    "    NUM_STEPS,  # Number of time steps\n",
+    "    NUM_SHOTS,  # Number of shots per time step\n",
+    "    output_dir,\n",
+    "    statevector_snapshots=True,\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "52a7ae28",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "qlbm-cpu-venv",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.13.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}