{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "QiGealjaVJNw"
   },
   "source": [
    "# Multi-Agent Plan Simple Example\n",
    "\n",
    "This python notebook shows how to use the unified planning library to model multi-agent problems.\n",
    "\n",
    "\n",
    "[![Open In GitHub](https://img.shields.io/badge/see-Github-579aca?logo=github)](https:///github.com/aiplan4eu/unified-planning/blob/master/docs/notebooks/09-multiagent-planning-simple.ipynb)\n",
    "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/aiplan4eu/unified-planning/blob/master/docs/notebooks/09-multiagent-planning-simple.ipynb)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "JvKlg7cm8tBo"
   },
   "source": [
    "## Setup\n",
    "\n",
    "We start by installing the unified planning library and a multi-agent planner (FMAP)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "aIxg95xz5EMo",
    "outputId": "44a3c006-f512-4f4e-9051-fab23c83970d",
    "tags": [
     "remove_from_CI"
    ]
   },
   "outputs": [],
   "source": [
    "!pip install unified-planning[fmap]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "OY-4EFALUj00"
   },
   "source": [
    "We are now ready to use the Unified-Planning library!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Vh5X0ZqRVVEA"
   },
   "source": [
    "## Demo\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "LY7O_siBU0V9"
   },
   "source": [
    "In this demo we show how to model a multi-agent planning problem using the Unified Planning library."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "tiQrjJOsV5zM"
   },
   "source": [
    "### Basic imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "dF8NSQydRRBO"
   },
   "outputs": [],
   "source": [
    "from unified_planning.shortcuts import *\n",
    "from unified_planning.model.multi_agent import *\n",
    "from collections import namedtuple\n",
    "from unified_planning.io.ma_pddl_writer import MAPDDLWriter"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "lUvelsbfWE9f"
   },
   "source": [
    "### Creating the simple-MA problem"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "WK2NwCw7XPB0"
   },
   "source": [
    "The class that represents a multi-agent planning problem is unified_planning.MultiAgentProblem, it contains the set of agents, the objects, an intial value for all the fluents and a goal to be reached by the planner.\n",
    "We create a `MultiAgentProblem` and two `Agents`. An `Agent` is an individual entity with specific fluents and specific\n",
    "actions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "xtPm63jkRRIE"
   },
   "outputs": [],
   "source": [
    "problem = MultiAgentProblem(\"simple_MA\")\n",
    "\n",
    "#AGENTs\n",
    "robot_a = Agent(\"robot_a\", problem)\n",
    "scale_a = Agent(\"scale_a\", problem)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "_4P9_BnPeXA-"
   },
   "source": [
    "We define the UserTypes, the Objects, the Fluents and the Actions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "id": "66JwC79NRRMz"
   },
   "outputs": [],
   "source": [
    "#USERTYPEs\n",
    "Location = UserType(\"Location\")\n",
    "door = UserType(\"door\")\n",
    "\n",
    "#FLUENTs\n",
    "open = Fluent(\"open\", door=door)\n",
    "pos = Fluent(\"pos\", loc=Location)\n",
    "\n",
    "#OBJECTs\n",
    "home = Object(\"home\", Location)\n",
    "office = Object(\"office\", Location)\n",
    "open20 = Object(\"open20\", door)\n",
    "close20 = Object(\"close20\", door)\n",
    "\n",
    "#ACTIONs\n",
    "movegripper = InstantaneousAction(\"movegripper\")\n",
    "movegripper.add_precondition(pos(office))\n",
    "movegripper.add_effect(pos(home), True)\n",
    "\n",
    "open_door = InstantaneousAction(\"open_door\")\n",
    "open_door.add_precondition(open(close20))\n",
    "open_door.add_effect(open(open20), True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "NIejvKXAetMK"
   },
   "source": [
    "Let's add the `Fluents` to the `Agents`. Note: An agent's fluent can be of two types, public or private. `Public Fluents` are visible to other agents. In contrast, the `Private Fluents` are not visible from the other agents.\n",
    "Private fluents are added to the agent via the `add_fluent` or `add_private_fluent` methods and public fluents via the `add_public_fluent` method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "LhlkeJ-ge2of",
    "outputId": "21cceb3b-a180-49cf-dff0-3a5a1e339a96"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bool open[door=door]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "robot_a.add_fluent(pos, default_initial_value=False)\n",
    "scale_a.add_fluent(open, default_initial_value=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4uhPijdDivoR"
   },
   "source": [
    "We add to the agents the actions they can perform."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "id": "4pCaQ6cQRRQc"
   },
   "outputs": [],
   "source": [
    "robot_a.add_action(movegripper)\n",
    "scale_a.add_action(open_door)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "WOFJtvHnjC_I"
   },
   "source": [
    "Let's add the agents to the `MultiAgentProblem`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "id": "ZK0kyCHUjCVi"
   },
   "outputs": [],
   "source": [
    "problem.add_agent(robot_a)\n",
    "problem.add_agent(scale_a)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Pl78ABIni9FV"
   },
   "source": [
    "We add the objects, the initial values and the goals. \n",
    "Note: `Dot` operator is used to denote agent-specific `Fluents`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "id": "tXu5TET8RRWx"
   },
   "outputs": [],
   "source": [
    "#OBJECTs\n",
    "problem.add_object(home)\n",
    "problem.add_object(office)\n",
    "problem.add_object(open20)\n",
    "problem.add_object(close20)\n",
    "\n",
    "#INITIAL VALUEs\n",
    "problem.set_initial_value(Dot(robot_a, pos(office)), True)\n",
    "problem.set_initial_value(Dot(scale_a, open(close20)), True)\n",
    "\n",
    "#GOALs\n",
    "problem.add_goal(Dot(robot_a, pos(home)))\n",
    "problem.add_goal(Dot(scale_a, open(open20)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "-mJ_Lrl6nTcQ"
   },
   "source": [
    "### MA-PDDL Writer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "KEkzMePBpB3x"
   },
   "source": [
    "To write the ma-pddl equivalent of a `unified_planning MultiAgentProblem` to a file we use the `MAPDDLWriter.write_ma_domain` and `MAPDDLWriter.write_ma_problem` methods."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "id": "zI9eG8iWRvUe"
   },
   "outputs": [],
   "source": [
    "w = MAPDDLWriter(problem)\n",
    "w.write_ma_domain(\"simple_ma\")\n",
    "w.write_ma_problem(\"simple_ma\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "UxrMTuOqVmzJ"
   },
   "source": [
    "## Solving Multi-Agent Planning Problems"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9SRCqarrVo6H"
   },
   "source": [
    "We can use fmap to solve a Multi-Agent problem.\n",
    "It allows the specification of a specific heuristic function used to evaluate the quality of the plans. The name of the custom parameter is heuristic and the following values are supported:\n",
    "\n",
    "  - 0 - FF heuristic: guides the search through the well-known h_FF heuristic function. This option is available for single-agent planning tasks only.\n",
    "  - 1 - DTG heuristic: evaluates plans via the heuristic h_DTG.\n",
    "  - 2 - default option - DTG + Landmarks: this option applies the multi-heuristic search scheme of the MH-FMAP solver by combining the h_DTG and h_Land heuristics to guide the search.\n",
    "  - 3 - Inc. DTG + Landmarks: incremental multi-heuristic mode that makes use of h_DTG and h_Land."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "4OF98T8i5ZOl",
    "outputId": "e812e0be-6c56-4903-f2ab-e4ddf331f289"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[96m\u001b[1mNOTE: To disable printing of planning engine credits, add this line to your code: `up.shortcuts.get_environment().credits_stream = None`\n",
      "\u001b[0m\u001b[96m  *** Credits ***\n",
      "\u001b[0m\u001b[96m  * In operation mode `OneshotPlanner` at line 1 of `/tmp/ipykernel_157727/678844819.py`, \u001b[0m\u001b[96myou are using the following planning engine:\n",
      "\u001b[0m\u001b[96m  * Engine name: FMAP\n",
      "  * Developers:  Alejandro Torreño, Oscar Sapena and Eva Onaindia\n",
      "\u001b[0m\u001b[96m  * Description: \u001b[0m\u001b[96mFMAP: A Platform for the Development of Distributed Multi-Agent Planning Systems.\u001b[0m\u001b[96m\n",
      "\u001b[0m\u001b[96m\n",
      "\u001b[0mFMAP Returned Sequential Plans object: <generator object PartialOrderPlan.all_sequential_plans at 0x7fec001b1d90>\n",
      "0 Sequential Plans: SequentialPlan:\n",
      "    scale_a.open_door\n",
      "    robot_a.movegripper\n",
      "1 Sequential Plans: SequentialPlan:\n",
      "    robot_a.movegripper\n",
      "    scale_a.open_door\n",
      "Adjacency list: {robot_a.movegripper: [], scale_a.open_door: []}\n",
      "result: status: SOLVED_SATISFICING\n",
      "engine: FMAP\n",
      "plan: PartialOrderPlan:\n",
      "  actions:\n",
      "    0) robot_a.movegripper\n",
      "    1) scale_a.open_door\n",
      "  constraints:\n"
     ]
    }
   ],
   "source": [
    "with OneshotPlanner(name='fmap', params={'heuristic': '1'}) as planner:\n",
    "    result = planner.solve(problem)\n",
    "    if result.status == up.engines.PlanGenerationResultStatus.SOLVED_SATISFICING:\n",
    "        print(\"%s Returned Sequential Plans object: %s\" % (planner.name, result.plan.all_sequential_plans()))\n",
    "        [print(f\"{idx} Sequential Plans: {seq_plan}\") for idx, seq_plan in enumerate(result.plan.all_sequential_plans())]\n",
    "        print(\"Adjacency list:\", result.plan.get_adjacency_list)\n",
    "        print(\"result:\", result)\n",
    "    else:\n",
    "        print(\"Log Error:\", result)"
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Tags",
  "colab": {
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}