Kopia_notatnika_[WSP]_Wfcommons.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "DXpM7fF5gN32",
   "metadata": {
    "id": "DXpM7fF5gN32"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "omCm46HZgPKi",
   "metadata": {
    "id": "omCm46HZgPKi"
   },
   "outputs": [],
   "source": [
    "!pip install -q pygraphviz"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "UihtvMz3f9UC",
   "metadata": {
    "id": "UihtvMz3f9UC"
   },
   "outputs": [],
   "source": [
    "!pip install -q wfcommons"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "41bc8a0a",
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install -q parse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "1a32b4a6",
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install -q dimod"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "62ed00c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install -q dwave-ocean-sdk\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "62f9a49f",
   "metadata": {
    "id": "62f9a49f",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "from wfcommons.wfchef.recipes import *\n",
    "from wfcommons.wfchef.utils import *\n",
    "from wfcommons import WorkflowGenerator\n",
    "import networkx as nx\n",
    "from networkx.drawing.nx_agraph import graphviz_layout\n",
    "import matplotlib.pyplot as plt\n",
    "import requests\n",
    "from pprint import pprint\n",
    "\n",
    "from typing import List"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ZSbP4M2tgopN",
   "metadata": {
    "id": "ZSbP4M2tgopN"
   },
   "source": [
    "# The first option is to generate a synthetic workflow based on some predefined patterns\n",
    "e.g., BlastRecipe, BwaRecipe, CyclesRecipe, EpigenomicsRecipe, GenomeRecipe, MontageRecipe, SeismologyRecipe, SoykbRecipe, SrasearchRecipe. See [here](https://wfcommons.org/generator) and [here](https://github.com/wfcommons/wfcommons/tree/438983117dc9ccf8b642339d81edef2ae992e34d/wfcommons/wfgen).\n",
    "\n",
    "-- issue: the machines field won't be filled (it provides only jobs)\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "4743c37b",
   "metadata": {
    "id": "4743c37b",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "# TASKS_NUM = 25\n",
    "# generator = WorkflowGenerator(SrasearchRecipe.from_num_tasks(TASKS_NUM))\n",
    "# workflow = generator.build_workflow()\n",
    "# DAG_synthetic = workflow.to_nx_digraph()\n",
    "# workflow.write_json('srasearch-workflow.json')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "jmhzVQ_47bMH",
   "metadata": {
    "id": "jmhzVQ_47bMH",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "HbGxVp0XkWku",
   "metadata": {
    "id": "HbGxVp0XkWku",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "def display_graph(graph):\n",
    "  pos = graphviz_layout(graph, prog='dot')\n",
    "  plt.figure(figsize=(14,10)) \n",
    "  nx.draw(graph, pos=pos, node_size=2000, labels ={\"SRC\" : \"Start\", \"DST\" : \"End\"}, with_labels=True, arrows=True)\n",
    "  plt.savefig(\"node_graph.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "XYG4dNQrkZ9g",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 751
    },
    "id": "XYG4dNQrkZ9g",
    "outputId": "17e526b4-0ccf-4e71-e11a-0fa884d6144b",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "# display_graph(DAG_synthetic)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "mqN8LjBchyYF",
   "metadata": {
    "id": "mqN8LjBchyYF"
   },
   "source": [
    "# The second option is to read a workflow from a file\n",
    "There are a few available workflow instances on Github: [pegasus-instances](https://github.com/wfcommons/pegasus-instances) and [makeflow-instances](https://github.com/wfcommons/makeflow-instances). There're json files where jobs are already assigned to machines."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b4acc22b",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "b4acc22b",
    "outputId": "0b56bcad-70b4-4e4b-b593-b2b65645c703",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "url = 'https://raw.githubusercontent.com/wfcommons/pegasus-instances/master/srasearch/chameleon-cloud/srasearch-chameleon-10a-001.json'\n",
    "# url = 'https://raw.githubusercontent.com/wfcommons/pegasus-instances/master/1000genome/chameleon-cloud/1000genome-chameleon-14ch-100k-001.json'\n",
    "resp = requests.get(url)\n",
    "data = json.loads(resp.text)\n",
    "pprint(data)\n",
    "path = 'pegasus-workflow.json'\n",
    "# workflow.write_json(path)\n",
    "with open(path, 'w') as f:\n",
    "    json.dump(data, f)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "3652576e",
   "metadata": {
    "id": "3652576e",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "DAG_real = create_graph(path)\n",
    "# annotate(DAG) # I thought that this line migth append some data to the DAG, but that's not the case https://github.com/wfcommons/wfcommons/blob/4abe2545998a7b29e65e94f95d7c32d7c9afdc78/wfcommons/wfchef/utils.py"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "800c0ab5",
   "metadata": {
    "id": "800c0ab5",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "nodes = DAG_real.nodes()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "6a612e2d",
   "metadata": {
    "id": "6a612e2d",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "edges = DAG_real.edges()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "0c394edb",
   "metadata": {
    "id": "0c394edb",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "in_degrees = DAG_real.in_degree(nodes)\n",
    "out_degrees = DAG_real.out_degree(nodes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "b5010194",
   "metadata": {
    "id": "b5010194",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "nodes_connected_to_source = [node for (node, in_degree) in in_degrees if in_degree == 0]\n",
    "nodes_connected_to_sink = [node for (node, out_degree) in out_degrees if out_degree == 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "2f390ada",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "2f390ada",
    "outputId": "9717930e-6f57-4362-ba21-14f44abea641",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nodes_connected_to_source ['SRC']\n",
      "nodes_connected_to_sink ['DST']\n"
     ]
    }
   ],
   "source": [
    "print('nodes_connected_to_source', nodes_connected_to_source)\n",
    "print('nodes_connected_to_sink', nodes_connected_to_sink)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "a2260590",
   "metadata": {
    "id": "a2260590",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "paths = list(nx.all_simple_paths(DAG_real, source='SRC', target='DST'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "1e117edb",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "1e117edb",
    "outputId": "953d3629-690b-4bad-d7c4-eb9e87389b4f",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Amount of paths: 20\n"
     ]
    }
   ],
   "source": [
    "paths_amount = len(paths)\n",
    "print(f'Amount of paths: {paths_amount}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "b64c2eb1",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 751
    },
    "id": "b64c2eb1",
    "outputId": "df599e34-af12-4785-aff6-f0cc22dd833b",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "module 'matplotlib.pyplot' has no attribute 'save'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[1;32m/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_[WSP]_Wfcommons.ipynb Cell 23'\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> <a href='vscode-notebook-cell://wsl%2Bubuntu/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_%5BWSP%5D_Wfcommons.ipynb#ch0000022vscode-remote?line=0'>1</a>\u001b[0m display_graph(DAG_real)\n",
      "\u001b[1;32m/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_[WSP]_Wfcommons.ipynb Cell 11'\u001b[0m in \u001b[0;36mdisplay_graph\u001b[0;34m(graph)\u001b[0m\n\u001b[1;32m      <a href='vscode-notebook-cell://wsl%2Bubuntu/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_%5BWSP%5D_Wfcommons.ipynb#ch0000010vscode-remote?line=2'>3</a>\u001b[0m plt\u001b[39m.\u001b[39mfigure(figsize\u001b[39m=\u001b[39m(\u001b[39m14\u001b[39m,\u001b[39m10\u001b[39m)) \n\u001b[1;32m      <a href='vscode-notebook-cell://wsl%2Bubuntu/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_%5BWSP%5D_Wfcommons.ipynb#ch0000010vscode-remote?line=3'>4</a>\u001b[0m nx\u001b[39m.\u001b[39mdraw(graph, pos\u001b[39m=\u001b[39mpos, node_size\u001b[39m=\u001b[39m\u001b[39m2000\u001b[39m, labels \u001b[39m=\u001b[39m{\u001b[39m\"\u001b[39m\u001b[39mSRC\u001b[39m\u001b[39m\"\u001b[39m : \u001b[39m\"\u001b[39m\u001b[39mStart\u001b[39m\u001b[39m\"\u001b[39m, \u001b[39m\"\u001b[39m\u001b[39mDST\u001b[39m\u001b[39m\"\u001b[39m : \u001b[39m\"\u001b[39m\u001b[39mEnd\u001b[39m\u001b[39m\"\u001b[39m}, with_labels\u001b[39m=\u001b[39m\u001b[39mTrue\u001b[39;00m, arrows\u001b[39m=\u001b[39m\u001b[39mTrue\u001b[39;00m)\n\u001b[0;32m----> <a href='vscode-notebook-cell://wsl%2Bubuntu/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_%5BWSP%5D_Wfcommons.ipynb#ch0000010vscode-remote?line=4'>5</a>\u001b[0m plt\u001b[39m.\u001b[39;49msave(\u001b[39m\"\u001b[39m\u001b[39mnode_graph.pdf\u001b[39m\u001b[39m\"\u001b[39m)\n",
      "\u001b[0;31mAttributeError\u001b[0m: module 'matplotlib.pyplot' has no attribute 'save'"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1008x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display_graph(DAG_real)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "R2URhNb4YVvl",
   "metadata": {
    "id": "R2URhNb4YVvl"
   },
   "source": [
    "So lets get some data :)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "FD7Z-KBfeEcD",
   "metadata": {
    "id": "FD7Z-KBfeEcD",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "class Machine(object):\n",
    "  def __init__(self, machine):\n",
    "    self.name = machine.get('nodeName')\n",
    "    self.system = machine.get('system')\n",
    "    self.architecture = machine.get('architecture')\n",
    "    self.release = machine.get('release')\n",
    "    self.cpu = machine.get('cpu')\n",
    "    self.memory = machine.get('memory')\n",
    "  def __repr__(self):\n",
    "    return str(vars(self))\n",
    "\n",
    "  def __str__(self):\n",
    "    return str(vars(self))\n",
    "\n",
    "\n",
    "class Job(object):\n",
    "  def __init__(self, id, name, children, parents, machine, inputs, outputs, runtime, used_memeory, normalized_runtime):\n",
    "    self.id = id\n",
    "    self.name = name\n",
    "    self.children = children\n",
    "    self.parents = parents\n",
    "    self.machine = machine\n",
    "    self.inputs = inputs\n",
    "    self.outputs = outputs\n",
    "    self.runtime = runtime\n",
    "    self.used_memeory = used_memeory\n",
    "    self.normalized_runtime = normalized_runtime\n",
    "\n",
    "\n",
    "  def find_and_set_parents(self, jobs):\n",
    "    parent_jobs = []\n",
    "    for j in jobs:\n",
    "      for p in self.parents:\n",
    "        if p == j.name:\n",
    "          parent_jobs.append(j)\n",
    "    self.parents = parent_jobs\n",
    "\n",
    "  def find_and_set_children(self, jobs):\n",
    "    children_jobs = []\n",
    "    for j in jobs:\n",
    "      for c in self.children:\n",
    "        if c == j.name:\n",
    "          children_jobs.append(j)\n",
    "    self.children = children_jobs\n",
    "\n",
    "\n",
    "  def __repr__(self):\n",
    "    return str(f\"['{self.name}']: on machine '{self.machine.name}' parent: '{[p.name for p in self.parents]}' children: '{[c.name for c in self.children]}'\" )\n",
    "\n",
    "  # def __str__(self):\n",
    "  #   return str(vars(self))\n",
    "\n",
    "class JobBuilder(object):\n",
    "  def __init__(self, job, *, machine_name = None):\n",
    "    self.id = job.get('id')\n",
    "    self.name = job.get('name')\n",
    "    self.category =  job.get('category') \n",
    "    self.children_names = job.get('children') or [] \n",
    "    self.parent_names = job.get('parents') or [] \n",
    "    self.commmand = job['command'].get('program') \n",
    "    self.arguments = job['command'].get('arguments')\n",
    "    files = job.get('files')\n",
    "    if files is None:\n",
    "      self.inputs = []\n",
    "      self.outputs = []\n",
    "    else: \n",
    "      self.inputs = [ (file.get('name'), file.get('size')) for file in files if file.get('link') == 'input']\n",
    "      self.outputs = [ (file.get('name'), file.get('size')) for file in files if file.get('link') == 'output']\n",
    "      \n",
    "    self.avg_cpu_usage = job.get('avgCPU')\n",
    "    self.used_memeory = job.get('memory')\n",
    "    self.priotrity = job.get( 'priority')\n",
    "    self.runtime = job.get('runtime')\n",
    "    self.type = job.get('type')\n",
    "    self.normalized_runtime = 1\n",
    "\n",
    "  def __repr__(self):\n",
    "    return str(vars(self))\n",
    "\n",
    "  def __str__(self):\n",
    "    return str(vars(self))\n",
    "\n",
    "  def set_machine(self, machine : Machine):\n",
    "    self.machine = machine\n",
    "\n",
    "  def build(self) -> Job:\n",
    "    return Job(id = self.id, name = self.name, children = self.children_names, parents = self.parent_names, machine = self.machine, inputs = self.inputs, outputs = self.outputs, runtime = self.runtime, used_memeory = self.used_memeory, normalized_runtime = self.normalized_runtime)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "JcyZpxrtj1s2",
   "metadata": {
    "id": "JcyZpxrtj1s2"
   },
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "dF0WHi4SjfE0",
   "metadata": {
    "id": "dF0WHi4SjfE0",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "json_data = data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "L1ciwEkGZw4F",
   "metadata": {
    "cellView": "code",
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "L1ciwEkGZw4F",
    "outputId": "5d976808-532f-4ab2-8993-f70b63ff6928",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'Machine' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_[WSP]_Wfcommons.ipynb Cell 28'\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      <a href='vscode-notebook-cell://wsl%2Bubuntu/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_%5BWSP%5D_Wfcommons.ipynb#ch0000027vscode-remote?line=0'>1</a>\u001b[0m \u001b[39m#@title Domyślny tekst tytułu\u001b[39;00m\n\u001b[0;32m----> <a href='vscode-notebook-cell://wsl%2Bubuntu/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_%5BWSP%5D_Wfcommons.ipynb#ch0000027vscode-remote?line=1'>2</a>\u001b[0m machines \u001b[39m=\u001b[39m  [Machine(machine) \u001b[39mfor\u001b[39;00m machine \u001b[39min\u001b[39;00m json_data[\u001b[39m'\u001b[39m\u001b[39mworkflow\u001b[39m\u001b[39m'\u001b[39m][\u001b[39m'\u001b[39m\u001b[39mmachines\u001b[39m\u001b[39m'\u001b[39m]]\n\u001b[1;32m      <a href='vscode-notebook-cell://wsl%2Bubuntu/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_%5BWSP%5D_Wfcommons.ipynb#ch0000027vscode-remote?line=2'>3</a>\u001b[0m machines \u001b[39m=\u001b[39m { machine\u001b[39m.\u001b[39mname : machine \u001b[39mfor\u001b[39;00m machine \u001b[39min\u001b[39;00m machines}\n\u001b[1;32m      <a href='vscode-notebook-cell://wsl%2Bubuntu/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_%5BWSP%5D_Wfcommons.ipynb#ch0000027vscode-remote?line=4'>5</a>\u001b[0m jobs_b \u001b[39m=\u001b[39m [ (JobBuilder(job), job[\u001b[39m'\u001b[39m\u001b[39mmachine\u001b[39m\u001b[39m'\u001b[39m]) \u001b[39mfor\u001b[39;00m job \u001b[39min\u001b[39;00m json_data[\u001b[39m'\u001b[39m\u001b[39mworkflow\u001b[39m\u001b[39m'\u001b[39m][\u001b[39m'\u001b[39m\u001b[39mtasks\u001b[39m\u001b[39m'\u001b[39m]]\n",
      "\u001b[1;32m/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_[WSP]_Wfcommons.ipynb Cell 28'\u001b[0m in \u001b[0;36m<listcomp>\u001b[0;34m(.0)\u001b[0m\n\u001b[1;32m      <a href='vscode-notebook-cell://wsl%2Bubuntu/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_%5BWSP%5D_Wfcommons.ipynb#ch0000027vscode-remote?line=0'>1</a>\u001b[0m \u001b[39m#@title Domyślny tekst tytułu\u001b[39;00m\n\u001b[0;32m----> <a href='vscode-notebook-cell://wsl%2Bubuntu/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_%5BWSP%5D_Wfcommons.ipynb#ch0000027vscode-remote?line=1'>2</a>\u001b[0m machines \u001b[39m=\u001b[39m  [Machine(machine) \u001b[39mfor\u001b[39;00m machine \u001b[39min\u001b[39;00m json_data[\u001b[39m'\u001b[39m\u001b[39mworkflow\u001b[39m\u001b[39m'\u001b[39m][\u001b[39m'\u001b[39m\u001b[39mmachines\u001b[39m\u001b[39m'\u001b[39m]]\n\u001b[1;32m      <a href='vscode-notebook-cell://wsl%2Bubuntu/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_%5BWSP%5D_Wfcommons.ipynb#ch0000027vscode-remote?line=2'>3</a>\u001b[0m machines \u001b[39m=\u001b[39m { machine\u001b[39m.\u001b[39mname : machine \u001b[39mfor\u001b[39;00m machine \u001b[39min\u001b[39;00m machines}\n\u001b[1;32m      <a href='vscode-notebook-cell://wsl%2Bubuntu/mnt/c/Users/mateu/Source/Magisterka/Kopia_notatnika_%5BWSP%5D_Wfcommons.ipynb#ch0000027vscode-remote?line=4'>5</a>\u001b[0m jobs_b \u001b[39m=\u001b[39m [ (JobBuilder(job), job[\u001b[39m'\u001b[39m\u001b[39mmachine\u001b[39m\u001b[39m'\u001b[39m]) \u001b[39mfor\u001b[39;00m job \u001b[39min\u001b[39;00m json_data[\u001b[39m'\u001b[39m\u001b[39mworkflow\u001b[39m\u001b[39m'\u001b[39m][\u001b[39m'\u001b[39m\u001b[39mtasks\u001b[39m\u001b[39m'\u001b[39m]]\n",
      "\u001b[0;31mNameError\u001b[0m: name 'Machine' is not defined"
     ]
    }
   ],
   "source": [
    "#@title Domyślny tekst tytułu\n",
    "machines =  [Machine(machine) for machine in json_data['workflow']['machines']]\n",
    "machines = { machine.name : machine for machine in machines}\n",
    "\n",
    "jobs_b = [ (JobBuilder(job), job['machine']) for job in json_data['workflow']['tasks']]\n",
    "\n",
    "pprint(machines)\n",
    "for job, machine_name in jobs_b:\n",
    "    job.set_machine(machines[machine_name])\n",
    "    job.normalized_runtime = job.runtime / job.machine.cpu['speed'] * job.machine.cpu['count']\n",
    "\n",
    "min_runtime = min(map(lambda x: x[0].normalized_runtime, jobs_b))\n",
    "\n",
    "for j,_ in jobs_b:\n",
    "  j.normalized_runtime /= min_runtime\n",
    "\n",
    "jobs = [job.build() for job, _ in jobs_b]\n",
    "\n",
    "\n",
    "\n",
    "for j in jobs:\n",
    "  j.find_and_set_parents(jobs)\n",
    "  j.find_and_set_children(jobs)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "H2gAe-iPDqGE",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "H2gAe-iPDqGE",
    "outputId": "832f60d1-c015-4264-fa30-32b7732dc7c9",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Jobs: 22\n",
      "Machines: 1\n"
     ]
    }
   ],
   "source": [
    "print(f'Jobs: {len(jobs)}')\n",
    "print(f'Machines: {len(machines)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "gpErdYl5eJWw",
   "metadata": {
    "id": "gpErdYl5eJWw"
   },
   "source": [
    "# Cyfronet Machines\n",
    "\n",
    "Lets define some machines that are available for AGH."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "iJkglYRJZ0wC",
   "metadata": {
    "id": "iJkglYRJZ0wC",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "base_cpu_hour_price = 0.08\n",
    "base_gpu_hour_price = 5.0\n",
    "gpu_speed_vs_cpu = 100\n",
    "\n",
    "cpus = {'Zeus': 0.25, 'Prometeus': 1 , 'Ares': 1.6}\n",
    "cpus_base_memory = {'Zeus': 2, 'Prometeus': 4 , 'Ares': 8}\n",
    "cpus_more_memory_factor = {'Zeus': 0.2, 'Prometeus': 0.40 , 'Ares': 0.80}\n",
    "\n",
    "\n",
    "gpus = {'Prometeus': 1 , 'Ares': 1.6}\n",
    "gpus_base_memory = {'Prometeus': 4 , 'Ares': 8}\n",
    "gpus_more_memory_factor = {'Prometeus': 0.40 , 'Ares': 0.80}\n",
    "\n",
    "\n",
    "cyfronet_machines = {}\n",
    "for cpu in cpus.keys():\n",
    "  cyfronet_machines[f\"{cpu}Cpu\"] = {\n",
    "      'base_price' :  base_cpu_hour_price,\n",
    "      'base_memory' : cpus_base_memory[cpu],\n",
    "      'memory_cost_multiplayer' : cpus_more_memory_factor[cpu],\n",
    "      'speed': cpus[cpu]\n",
    "  }\n",
    "for gpu in gpus.keys():\n",
    "  cyfronet_machines[f\"{gpu}Gpu\"] = {\n",
    "      'base_price' : base_gpu_hour_price,\n",
    "      'base_memory' : gpus_base_memory[gpu],\n",
    "      'memory_cost_multiplayer' : gpus_more_memory_factor[gpu],\n",
    "      'speed' : gpus[gpu] * gpu_speed_vs_cpu\n",
    "  }\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "EeFL5UAHhj0K",
   "metadata": {
    "id": "EeFL5UAHhj0K",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "import math\n",
    "def calc_cost(machine , job):\n",
    "  if job.used_memeory is not None:\n",
    "    mib_free = 1024 * machine['base_memory']\n",
    "    cost_power = (job.used_memeory - mib_free) // 1024\n",
    "  else:\n",
    "    cost_power = 0\n",
    "  runtime = job.normalized_runtime\n",
    "  return runtime * machine['base_price'] * (1 + machine['memory_cost_multiplayer']* cost_power) / machine['speed']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "9_Sc0EeohkBI",
   "metadata": {
    "id": "9_Sc0EeohkBI",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "costs = {}\n",
    "runtimes = {}\n",
    "for machine in cyfronet_machines:\n",
    "  machine_cost = []\n",
    "  machine_runtime = []\n",
    "  for j in jobs:\n",
    "    machine_cost.append(calc_cost(cyfronet_machines[machine], j))\n",
    "    machine_runtime.append( j.normalized_runtime / cyfronet_machines[machine]['speed'])\n",
    "  costs[machine] = machine_cost\n",
    "  runtimes[f\"{machine}Runtime\"] = machine_runtime"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "ZbE8iPJDjJyJ",
   "metadata": {
    "id": "ZbE8iPJDjJyJ",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "WV3Nd7_-oOG2",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 739
    },
    "id": "WV3Nd7_-oOG2",
    "outputId": "019ba39d-fcde-43b0-e064-c287ae2f183d",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ZeusCpu</th>\n",
       "      <th>PrometeusCpu</th>\n",
       "      <th>AresCpu</th>\n",
       "      <th>PrometeusGpu</th>\n",
       "      <th>AresGpu</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>bowtie2-build_ID0000001</th>\n",
       "      <td>335.827478</td>\n",
       "      <td>159.959930</td>\n",
       "      <td>188.350609</td>\n",
       "      <td>99.974957</td>\n",
       "      <td>117.719130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000002</th>\n",
       "      <td>159221.188452</td>\n",
       "      <td>78487.560765</td>\n",
       "      <td>96471.693826</td>\n",
       "      <td>49054.725478</td>\n",
       "      <td>60294.808641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000003</th>\n",
       "      <td>873.724104</td>\n",
       "      <td>369.073113</td>\n",
       "      <td>362.482522</td>\n",
       "      <td>230.670696</td>\n",
       "      <td>226.551576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000004</th>\n",
       "      <td>46585.664000</td>\n",
       "      <td>22726.249600</td>\n",
       "      <td>27581.546000</td>\n",
       "      <td>14203.906000</td>\n",
       "      <td>17238.466250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000005</th>\n",
       "      <td>782.037704</td>\n",
       "      <td>330.343513</td>\n",
       "      <td>324.444522</td>\n",
       "      <td>206.464696</td>\n",
       "      <td>202.777826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000006</th>\n",
       "      <td>87037.907478</td>\n",
       "      <td>42741.829565</td>\n",
       "      <td>52293.980870</td>\n",
       "      <td>26713.643478</td>\n",
       "      <td>32683.738043</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000007</th>\n",
       "      <td>860.093217</td>\n",
       "      <td>358.372174</td>\n",
       "      <td>343.440000</td>\n",
       "      <td>223.982609</td>\n",
       "      <td>214.650000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000008</th>\n",
       "      <td>39946.646817</td>\n",
       "      <td>19447.709635</td>\n",
       "      <td>23543.116957</td>\n",
       "      <td>12154.818522</td>\n",
       "      <td>14714.448098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000009</th>\n",
       "      <td>673.529322</td>\n",
       "      <td>280.637217</td>\n",
       "      <td>268.944000</td>\n",
       "      <td>175.398261</td>\n",
       "      <td>168.090000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000010</th>\n",
       "      <td>114676.087096</td>\n",
       "      <td>56526.656139</td>\n",
       "      <td>69475.046870</td>\n",
       "      <td>35329.160087</td>\n",
       "      <td>43421.904293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000011</th>\n",
       "      <td>651.335791</td>\n",
       "      <td>271.389913</td>\n",
       "      <td>260.082000</td>\n",
       "      <td>169.618696</td>\n",
       "      <td>162.551250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000012</th>\n",
       "      <td>68406.293148</td>\n",
       "      <td>33522.110470</td>\n",
       "      <td>40909.460435</td>\n",
       "      <td>20951.319043</td>\n",
       "      <td>25568.412772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000013</th>\n",
       "      <td>714.670748</td>\n",
       "      <td>297.779478</td>\n",
       "      <td>285.372000</td>\n",
       "      <td>186.112174</td>\n",
       "      <td>178.357500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000014</th>\n",
       "      <td>35153.921113</td>\n",
       "      <td>17079.499409</td>\n",
       "      <td>20623.910087</td>\n",
       "      <td>10674.687130</td>\n",
       "      <td>12889.943804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000015</th>\n",
       "      <td>479.377252</td>\n",
       "      <td>199.740522</td>\n",
       "      <td>191.418000</td>\n",
       "      <td>124.837826</td>\n",
       "      <td>119.636250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000016</th>\n",
       "      <td>149033.959513</td>\n",
       "      <td>73465.799791</td>\n",
       "      <td>90299.278957</td>\n",
       "      <td>45916.124870</td>\n",
       "      <td>56437.049348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000017</th>\n",
       "      <td>1016.634991</td>\n",
       "      <td>423.597913</td>\n",
       "      <td>405.948000</td>\n",
       "      <td>264.748696</td>\n",
       "      <td>253.717500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000018</th>\n",
       "      <td>117514.342400</td>\n",
       "      <td>57872.865948</td>\n",
       "      <td>71051.470609</td>\n",
       "      <td>36170.541217</td>\n",
       "      <td>44407.169130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000019</th>\n",
       "      <td>974.757287</td>\n",
       "      <td>406.148870</td>\n",
       "      <td>389.226000</td>\n",
       "      <td>253.843043</td>\n",
       "      <td>243.266250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000020</th>\n",
       "      <td>164060.827826</td>\n",
       "      <td>80876.861217</td>\n",
       "      <td>99413.812174</td>\n",
       "      <td>50548.038261</td>\n",
       "      <td>62133.632609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000021</th>\n",
       "      <td>1363.804939</td>\n",
       "      <td>576.090017</td>\n",
       "      <td>565.802696</td>\n",
       "      <td>360.056261</td>\n",
       "      <td>353.626685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>merge_ID0000022</th>\n",
       "      <td>0.384000</td>\n",
       "      <td>0.048000</td>\n",
       "      <td>-0.150000</td>\n",
       "      <td>0.030000</td>\n",
       "      <td>-0.093750</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                               ZeusCpu  PrometeusCpu       AresCpu  \\\n",
       "bowtie2-build_ID0000001     335.827478    159.959930    188.350609   \n",
       "fasterq-dump_ID0000002   159221.188452  78487.560765  96471.693826   \n",
       "bowtie2_ID0000003           873.724104    369.073113    362.482522   \n",
       "fasterq-dump_ID0000004    46585.664000  22726.249600  27581.546000   \n",
       "bowtie2_ID0000005           782.037704    330.343513    324.444522   \n",
       "fasterq-dump_ID0000006    87037.907478  42741.829565  52293.980870   \n",
       "bowtie2_ID0000007           860.093217    358.372174    343.440000   \n",
       "fasterq-dump_ID0000008    39946.646817  19447.709635  23543.116957   \n",
       "bowtie2_ID0000009           673.529322    280.637217    268.944000   \n",
       "fasterq-dump_ID0000010   114676.087096  56526.656139  69475.046870   \n",
       "bowtie2_ID0000011           651.335791    271.389913    260.082000   \n",
       "fasterq-dump_ID0000012    68406.293148  33522.110470  40909.460435   \n",
       "bowtie2_ID0000013           714.670748    297.779478    285.372000   \n",
       "fasterq-dump_ID0000014    35153.921113  17079.499409  20623.910087   \n",
       "bowtie2_ID0000015           479.377252    199.740522    191.418000   \n",
       "fasterq-dump_ID0000016   149033.959513  73465.799791  90299.278957   \n",
       "bowtie2_ID0000017          1016.634991    423.597913    405.948000   \n",
       "fasterq-dump_ID0000018   117514.342400  57872.865948  71051.470609   \n",
       "bowtie2_ID0000019           974.757287    406.148870    389.226000   \n",
       "fasterq-dump_ID0000020   164060.827826  80876.861217  99413.812174   \n",
       "bowtie2_ID0000021          1363.804939    576.090017    565.802696   \n",
       "merge_ID0000022               0.384000      0.048000     -0.150000   \n",
       "\n",
       "                         PrometeusGpu       AresGpu  \n",
       "bowtie2-build_ID0000001     99.974957    117.719130  \n",
       "fasterq-dump_ID0000002   49054.725478  60294.808641  \n",
       "bowtie2_ID0000003          230.670696    226.551576  \n",
       "fasterq-dump_ID0000004   14203.906000  17238.466250  \n",
       "bowtie2_ID0000005          206.464696    202.777826  \n",
       "fasterq-dump_ID0000006   26713.643478  32683.738043  \n",
       "bowtie2_ID0000007          223.982609    214.650000  \n",
       "fasterq-dump_ID0000008   12154.818522  14714.448098  \n",
       "bowtie2_ID0000009          175.398261    168.090000  \n",
       "fasterq-dump_ID0000010   35329.160087  43421.904293  \n",
       "bowtie2_ID0000011          169.618696    162.551250  \n",
       "fasterq-dump_ID0000012   20951.319043  25568.412772  \n",
       "bowtie2_ID0000013          186.112174    178.357500  \n",
       "fasterq-dump_ID0000014   10674.687130  12889.943804  \n",
       "bowtie2_ID0000015          124.837826    119.636250  \n",
       "fasterq-dump_ID0000016   45916.124870  56437.049348  \n",
       "bowtie2_ID0000017          264.748696    253.717500  \n",
       "fasterq-dump_ID0000018   36170.541217  44407.169130  \n",
       "bowtie2_ID0000019          253.843043    243.266250  \n",
       "fasterq-dump_ID0000020   50548.038261  62133.632609  \n",
       "bowtie2_ID0000021          360.056261    353.626685  \n",
       "merge_ID0000022              0.030000     -0.093750  "
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cost_df = pd.DataFrame(data=costs, index = [j.name for j in jobs])\n",
    "cost_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "-h1x8j7TH5Ho",
   "metadata": {
    "id": "-h1x8j7TH5Ho",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "OZaZrqh6pz_t",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 739
    },
    "id": "OZaZrqh6pz_t",
    "outputId": "f027ee49-5cf2-459a-db01-eead99605993",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ZeusCpuRuntime</th>\n",
       "      <th>PrometeusCpuRuntime</th>\n",
       "      <th>AresCpuRuntime</th>\n",
       "      <th>PrometeusGpuRuntime</th>\n",
       "      <th>AresGpuRuntime</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>bowtie2-build_ID0000001</th>\n",
       "      <td>220.939130</td>\n",
       "      <td>55.234783</td>\n",
       "      <td>34.521739</td>\n",
       "      <td>0.552348</td>\n",
       "      <td>0.345217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000002</th>\n",
       "      <td>31195.373913</td>\n",
       "      <td>7798.843478</td>\n",
       "      <td>4874.277174</td>\n",
       "      <td>77.988435</td>\n",
       "      <td>48.742772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000003</th>\n",
       "      <td>1883.026087</td>\n",
       "      <td>470.756522</td>\n",
       "      <td>294.222826</td>\n",
       "      <td>4.707565</td>\n",
       "      <td>2.942228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000004</th>\n",
       "      <td>15738.400000</td>\n",
       "      <td>3934.600000</td>\n",
       "      <td>2459.125000</td>\n",
       "      <td>39.346000</td>\n",
       "      <td>24.591250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000005</th>\n",
       "      <td>1685.426087</td>\n",
       "      <td>421.356522</td>\n",
       "      <td>263.347826</td>\n",
       "      <td>4.213565</td>\n",
       "      <td>2.633478</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000006</th>\n",
       "      <td>21586.782609</td>\n",
       "      <td>5396.695652</td>\n",
       "      <td>3372.934783</td>\n",
       "      <td>53.966957</td>\n",
       "      <td>33.729348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000007</th>\n",
       "      <td>1990.956522</td>\n",
       "      <td>497.739130</td>\n",
       "      <td>311.086957</td>\n",
       "      <td>4.977391</td>\n",
       "      <td>3.110870</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000008</th>\n",
       "      <td>14600.382609</td>\n",
       "      <td>3650.095652</td>\n",
       "      <td>2281.309783</td>\n",
       "      <td>36.500957</td>\n",
       "      <td>22.813098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000009</th>\n",
       "      <td>1559.095652</td>\n",
       "      <td>389.773913</td>\n",
       "      <td>243.608696</td>\n",
       "      <td>3.897739</td>\n",
       "      <td>2.436087</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000010</th>\n",
       "      <td>22538.539130</td>\n",
       "      <td>5634.634783</td>\n",
       "      <td>3521.646739</td>\n",
       "      <td>56.346348</td>\n",
       "      <td>35.216467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000011</th>\n",
       "      <td>1507.721739</td>\n",
       "      <td>376.930435</td>\n",
       "      <td>235.581522</td>\n",
       "      <td>3.769304</td>\n",
       "      <td>2.355815</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000012</th>\n",
       "      <td>18917.669565</td>\n",
       "      <td>4729.417391</td>\n",
       "      <td>2955.885870</td>\n",
       "      <td>47.294174</td>\n",
       "      <td>29.558859</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000013</th>\n",
       "      <td>1654.330435</td>\n",
       "      <td>413.582609</td>\n",
       "      <td>258.489130</td>\n",
       "      <td>4.135826</td>\n",
       "      <td>2.584891</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000014</th>\n",
       "      <td>13818.365217</td>\n",
       "      <td>3454.591304</td>\n",
       "      <td>2159.119565</td>\n",
       "      <td>34.545913</td>\n",
       "      <td>21.591196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000015</th>\n",
       "      <td>1109.669565</td>\n",
       "      <td>277.417391</td>\n",
       "      <td>173.385870</td>\n",
       "      <td>2.774174</td>\n",
       "      <td>1.733859</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000016</th>\n",
       "      <td>29199.443478</td>\n",
       "      <td>7299.860870</td>\n",
       "      <td>4562.413043</td>\n",
       "      <td>72.998609</td>\n",
       "      <td>45.624130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000017</th>\n",
       "      <td>2353.321739</td>\n",
       "      <td>588.330435</td>\n",
       "      <td>367.706522</td>\n",
       "      <td>5.883304</td>\n",
       "      <td>3.677065</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000018</th>\n",
       "      <td>24564.034783</td>\n",
       "      <td>6141.008696</td>\n",
       "      <td>3838.130435</td>\n",
       "      <td>61.410087</td>\n",
       "      <td>38.381304</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000019</th>\n",
       "      <td>2256.382609</td>\n",
       "      <td>564.095652</td>\n",
       "      <td>352.559783</td>\n",
       "      <td>5.640957</td>\n",
       "      <td>3.525598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fasterq-dump_ID0000020</th>\n",
       "      <td>32043.130435</td>\n",
       "      <td>8010.782609</td>\n",
       "      <td>5006.739130</td>\n",
       "      <td>80.107826</td>\n",
       "      <td>50.067391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bowtie2_ID0000021</th>\n",
       "      <td>2939.234783</td>\n",
       "      <td>734.808696</td>\n",
       "      <td>459.255435</td>\n",
       "      <td>7.348087</td>\n",
       "      <td>4.592554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>merge_ID0000022</th>\n",
       "      <td>4.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.625000</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.006250</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         ZeusCpuRuntime  PrometeusCpuRuntime  AresCpuRuntime  \\\n",
       "bowtie2-build_ID0000001      220.939130            55.234783       34.521739   \n",
       "fasterq-dump_ID0000002     31195.373913          7798.843478     4874.277174   \n",
       "bowtie2_ID0000003           1883.026087           470.756522      294.222826   \n",
       "fasterq-dump_ID0000004     15738.400000          3934.600000     2459.125000   \n",
       "bowtie2_ID0000005           1685.426087           421.356522      263.347826   \n",
       "fasterq-dump_ID0000006     21586.782609          5396.695652     3372.934783   \n",
       "bowtie2_ID0000007           1990.956522           497.739130      311.086957   \n",
       "fasterq-dump_ID0000008     14600.382609          3650.095652     2281.309783   \n",
       "bowtie2_ID0000009           1559.095652           389.773913      243.608696   \n",
       "fasterq-dump_ID0000010     22538.539130          5634.634783     3521.646739   \n",
       "bowtie2_ID0000011           1507.721739           376.930435      235.581522   \n",
       "fasterq-dump_ID0000012     18917.669565          4729.417391     2955.885870   \n",
       "bowtie2_ID0000013           1654.330435           413.582609      258.489130   \n",
       "fasterq-dump_ID0000014     13818.365217          3454.591304     2159.119565   \n",
       "bowtie2_ID0000015           1109.669565           277.417391      173.385870   \n",
       "fasterq-dump_ID0000016     29199.443478          7299.860870     4562.413043   \n",
       "bowtie2_ID0000017           2353.321739           588.330435      367.706522   \n",
       "fasterq-dump_ID0000018     24564.034783          6141.008696     3838.130435   \n",
       "bowtie2_ID0000019           2256.382609           564.095652      352.559783   \n",
       "fasterq-dump_ID0000020     32043.130435          8010.782609     5006.739130   \n",
       "bowtie2_ID0000021           2939.234783           734.808696      459.255435   \n",
       "merge_ID0000022                4.000000             1.000000        0.625000   \n",
       "\n",
       "                         PrometeusGpuRuntime  AresGpuRuntime  \n",
       "bowtie2-build_ID0000001             0.552348        0.345217  \n",
       "fasterq-dump_ID0000002             77.988435       48.742772  \n",
       "bowtie2_ID0000003                   4.707565        2.942228  \n",
       "fasterq-dump_ID0000004             39.346000       24.591250  \n",
       "bowtie2_ID0000005                   4.213565        2.633478  \n",
       "fasterq-dump_ID0000006             53.966957       33.729348  \n",
       "bowtie2_ID0000007                   4.977391        3.110870  \n",
       "fasterq-dump_ID0000008             36.500957       22.813098  \n",
       "bowtie2_ID0000009                   3.897739        2.436087  \n",
       "fasterq-dump_ID0000010             56.346348       35.216467  \n",
       "bowtie2_ID0000011                   3.769304        2.355815  \n",
       "fasterq-dump_ID0000012             47.294174       29.558859  \n",
       "bowtie2_ID0000013                   4.135826        2.584891  \n",
       "fasterq-dump_ID0000014             34.545913       21.591196  \n",
       "bowtie2_ID0000015                   2.774174        1.733859  \n",
       "fasterq-dump_ID0000016             72.998609       45.624130  \n",
       "bowtie2_ID0000017                   5.883304        3.677065  \n",
       "fasterq-dump_ID0000018             61.410087       38.381304  \n",
       "bowtie2_ID0000019                   5.640957        3.525598  \n",
       "fasterq-dump_ID0000020             80.107826       50.067391  \n",
       "bowtie2_ID0000021                   7.348087        4.592554  \n",
       "merge_ID0000022                     0.010000        0.006250  "
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "runtime_df = pd.DataFrame(data=runtimes, index = [j.name for j in jobs])\n",
    "runtime_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "jpUeBNdOw-_v",
   "metadata": {
    "id": "jpUeBNdOw-_v",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "UZJxsI_gxU_V",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "UZJxsI_gxU_V",
    "outputId": "f0baccfc-b285-45d8-8e9c-8c095ffcefc7",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Shape (110, 20)'\n",
      "array([[2.2093913e+02, 2.2093913e+02, 2.2093913e+02, ..., 0.0000000e+00,\n",
      "        0.0000000e+00, 0.0000000e+00],\n",
      "       [0.0000000e+00, 0.0000000e+00, 0.0000000e+00, ..., 0.0000000e+00,\n",
      "        0.0000000e+00, 0.0000000e+00],\n",
      "       [1.8830261e+03, 0.0000000e+00, 0.0000000e+00, ..., 0.0000000e+00,\n",
      "        0.0000000e+00, 0.0000000e+00],\n",
      "       ...,\n",
      "       [0.0000000e+00, 0.0000000e+00, 0.0000000e+00, ..., 0.0000000e+00,\n",
      "        0.0000000e+00, 5.0067390e+01],\n",
      "       [0.0000000e+00, 0.0000000e+00, 0.0000000e+00, ..., 0.0000000e+00,\n",
      "        0.0000000e+00, 4.5925546e+00],\n",
      "       [6.2500001e-03, 6.2500001e-03, 6.2500001e-03, ..., 6.2500001e-03,\n",
      "        6.2500001e-03, 6.2500001e-03]], dtype=float32)\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# flat_runtimes = [runtime for n, machine_runtimes in runtimes.items() for runtime in machine_runtimes]\n",
    "flat_runtimes = [(runtime, name) for n, machine_runtimes in runtimes.items() for runtime, name in zip(machine_runtimes, [j.name for j in jobs])]\n",
    "\n",
    "R = np.zeros(shape=(len(flat_runtimes),len(paths)), dtype=np.float32)\n",
    "\n",
    "for i, (job, name) in enumerate(flat_runtimes):\n",
    "    for j, path in enumerate(paths):\n",
    "      for node in path:\n",
    "        if name == node:\n",
    "          R[i,j] = job\n",
    "\n",
    "pprint(f'Shape {R.shape}')\n",
    "pprint(R)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "MHMaJtkMyGK3",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "MHMaJtkMyGK3",
    "outputId": "98b03ec8-908f-43ac-9413-dbd488fea97a",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 1., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 1., ..., 0., 0., 0.],\n",
       "       ...,\n",
       "       [0., 0., 0., ..., 1., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 1., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 1.]], dtype=float32)"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_size =  np.array(cost_df).shape[0] * np.array(cost_df).shape[1]\n",
    "USE_ONE_MACHINE  = np.zeros(shape=( y_size, np.array(cost_df).shape[0]), dtype = np.float32)\n",
    "for i in range(len(jobs)):\n",
    "  for j in range(i, y_size, len(jobs)):\n",
    "    USE_ONE_MACHINE[j,i] = 1\n",
    "\n",
    "USE_ONE_MACHINE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "E3uM5d1JT3M_",
   "metadata": {
    "id": "E3uM5d1JT3M_",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "DEADLINE = 40000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "qYZoqaKeT3XP",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "qYZoqaKeT3XP",
    "outputId": "05aeba06-3bf2-423e-a889-c316f7b74ee5",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 220,  220,  220, ...,    0,    0,    0],\n",
       "       [   0,    0,    0, ...,    0,    0,    0],\n",
       "       [1883,    0,    0, ...,    0,    0,    0],\n",
       "       ...,\n",
       "       [   0,    0,    0, ...,    0,    0,   50],\n",
       "       [   0,    0,    0, ...,    0,    0,    4],\n",
       "       [   0,    0,    0, ...,    0,    0,    0]], dtype=int32)"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "PATHS =  (R).astype(np.int32)\n",
    "PATHS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "gmko9CeiUaoR",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "gmko9CeiUaoR",
    "outputId": "dee17514-e79c-49ae-d5ff-483aeedd7a0c",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[    1,     2,     4, ...,     0,     0,     0],\n",
       "       [    0,     0,     0, ...,     0,     0,     0],\n",
       "       [    0,     0,     0, ...,     0,     0,     0],\n",
       "       ...,\n",
       "       [    0,     0,     0, ...,     0,     0,     0],\n",
       "       [    0,     0,     0, ...,     0,     0,     0],\n",
       "       [    0,     0,     0, ...,  8192, 16384, 32768]], dtype=int32)"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "POWERS_2 = [ 2 ** i for i in range(np.ceil(np.log2(DEADLINE)).astype(np.int32))]\n",
    "POWERS_2\n",
    "\n",
    "SLACK = np.zeros(shape=(PATHS.shape[0], PATHS.shape[0] * len(POWERS_2))).astype(np.int32)\n",
    "\n",
    "for i in range(0, SLACK.shape[0]):\n",
    "    SLACK[i, i * len(POWERS_2):(i + 1) * len(POWERS_2)] = POWERS_2\n",
    "SLACK"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "U-CTkNB1Wvyf",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "U-CTkNB1Wvyf",
    "outputId": "326f595f-0eb5-42dc-d56c-2560556d87b4",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PATHS shape (110, 20)\n",
      "SLACK shape (110, 1760)\n",
      "USE_ONE_MACHINE shape (110, 22)\n",
      "Intermediate A shape (110, 1782)\n",
      "U shape (110, 1782)\n",
      "===================================================\n",
      "A shape (220, 1782)\n",
      "b shape (220,)\n"
     ]
    }
   ],
   "source": [
    "A = np.concatenate((PATHS,SLACK), axis=1)\n",
    "\n",
    "\n",
    "U = np.concatenate((USE_ONE_MACHINE, np.zeros(shape = (USE_ONE_MACHINE.shape[0], SLACK.shape[1]))), axis = 1)\n",
    "if U.shape[1] - A.shape[1] > 0:\n",
    "  A = np.pad(A, [(0, 0), (0, U.shape[1] - A.shape[1])], mode='constant', constant_values=0)\n",
    "elif U.shape[1] - A.shape[1] < 0:\n",
    "  U = np.pad(U, [(0, 0), (0, A.shape[1] - U.shape[1])], mode='constant', constant_values=0)\n",
    "\n",
    "b_paths = np.array([DEADLINE for i in range(A.shape[0])] , dtype=np.int32 )\n",
    "b_machines = np.array([1 for i in range(U.shape[0]) ] , dtype=np.int32 )\n",
    "b = np.concatenate((b_paths, b_machines), axis = 0)\n",
    "  \n",
    "A = np.pad(A, [(0, 0), (0, U.shape[1] - A.shape[1])], mode='constant', constant_values=0)\n",
    "S = 10000\n",
    "\n",
    "print(f'PATHS shape {PATHS.shape}')\n",
    "print(f'SLACK shape {SLACK.shape}')\n",
    "print(f'USE_ONE_MACHINE shape {USE_ONE_MACHINE.shape}')\n",
    "\n",
    "print(f'Intermediate A shape {A.shape}')\n",
    "print(f'U shape {U.shape}')\n",
    "\n",
    "A = np.concatenate((A, S * U))\n",
    "\n",
    "print(f'===================================================')\n",
    "print(f'A shape {A.shape}')\n",
    "print(f'b shape {b.shape}')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "VSFmKgZEalSt",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "VSFmKgZEalSt",
    "outputId": "25353b07-b796-487e-d9b8-156357f6ffa3",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "C shape (110,)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([ 3.35827478e+02,  1.59959930e+02,  1.88350609e+02,  9.99749565e+01,\n",
       "        1.17719130e+02,  1.59221188e+05,  7.84875608e+04,  9.64716938e+04,\n",
       "        4.90547255e+04,  6.02948086e+04,  8.73724104e+02,  3.69073113e+02,\n",
       "        3.62482522e+02,  2.30670696e+02,  2.26551576e+02,  4.65856640e+04,\n",
       "        2.27262496e+04,  2.75815460e+04,  1.42039060e+04,  1.72384663e+04,\n",
       "        7.82037704e+02,  3.30343513e+02,  3.24444522e+02,  2.06464696e+02,\n",
       "        2.02777826e+02,  8.70379075e+04,  4.27418296e+04,  5.22939809e+04,\n",
       "        2.67136435e+04,  3.26837380e+04,  8.60093217e+02,  3.58372174e+02,\n",
       "        3.43440000e+02,  2.23982609e+02,  2.14650000e+02,  3.99466468e+04,\n",
       "        1.94477096e+04,  2.35431170e+04,  1.21548185e+04,  1.47144481e+04,\n",
       "        6.73529322e+02,  2.80637217e+02,  2.68944000e+02,  1.75398261e+02,\n",
       "        1.68090000e+02,  1.14676087e+05,  5.65266561e+04,  6.94750469e+04,\n",
       "        3.53291601e+04,  4.34219043e+04,  6.51335791e+02,  2.71389913e+02,\n",
       "        2.60082000e+02,  1.69618696e+02,  1.62551250e+02,  6.84062931e+04,\n",
       "        3.35221105e+04,  4.09094604e+04,  2.09513190e+04,  2.55684128e+04,\n",
       "        7.14670748e+02,  2.97779478e+02,  2.85372000e+02,  1.86112174e+02,\n",
       "        1.78357500e+02,  3.51539211e+04,  1.70794994e+04,  2.06239101e+04,\n",
       "        1.06746871e+04,  1.28899438e+04,  4.79377252e+02,  1.99740522e+02,\n",
       "        1.91418000e+02,  1.24837826e+02,  1.19636250e+02,  1.49033960e+05,\n",
       "        7.34657998e+04,  9.02992790e+04,  4.59161249e+04,  5.64370493e+04,\n",
       "        1.01663499e+03,  4.23597913e+02,  4.05948000e+02,  2.64748696e+02,\n",
       "        2.53717500e+02,  1.17514342e+05,  5.78728659e+04,  7.10514706e+04,\n",
       "        3.61705412e+04,  4.44071691e+04,  9.74757287e+02,  4.06148870e+02,\n",
       "        3.89226000e+02,  2.53843043e+02,  2.43266250e+02,  1.64060828e+05,\n",
       "        8.08768612e+04,  9.94138122e+04,  5.05480383e+04,  6.21336326e+04,\n",
       "        1.36380494e+03,  5.76090017e+02,  5.65802696e+02,  3.60056261e+02,\n",
       "        3.53626685e+02,  3.84000000e-01,  4.80000000e-02, -1.50000000e-01,\n",
       "        3.00000000e-02, -9.37500000e-02])"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "C_diag = np.diag(np.array(cost_df).flatten())\n",
    "C = np.array(cost_df).flatten()\n",
    "print(f'C shape {C.shape}')\n",
    "C"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "3hCbRFVFAkZX",
   "metadata": {
    "id": "3hCbRFVFAkZX",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "P = 10000\n",
    "\n",
    "# SOLUTION\n",
    "D = np.diag(2 * A.transpose().dot(b))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "9TdkAY7dwl73",
   "metadata": {
    "id": "9TdkAY7dwl73",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "C = np.pad(C, [(0, D.shape[0] - C.shape[0])], mode='constant')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "Npm2EyNvwj4U",
   "metadata": {
    "id": "Npm2EyNvwj4U",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "QUBO = P * (A.transpose().dot(A) + D) + C"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "383kXyjOxGDI",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "383kXyjOxGDI",
    "outputId": "a7ce21ea-b46d-45f0-8e18-d301c8632c1b",
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[7.41365643e+12, 5.25980160e+08, 5.25980188e+08, ...,\n",
       "        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
       "       [5.25980336e+08, 7.18358253e+12, 5.25980188e+08, ...,\n",
       "        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
       "       [5.25980336e+08, 5.25980160e+08, 7.53976453e+12, ...,\n",
       "        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
       "       ...,\n",
       "       [3.35827478e+02, 1.59959930e+02, 1.88350609e+02, ...,\n",
       "        3.69518182e+13, 0.00000000e+00, 0.00000000e+00],\n",
       "       [3.35827478e+02, 1.59959930e+02, 1.88350609e+02, ...,\n",
       "        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
       "       [3.35827478e+02, 1.59959930e+02, 1.88350609e+02, ...,\n",
       "        0.00000000e+00, 0.00000000e+00, 0.00000000e+00]])"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "QUBO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "4531719f",
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "import dimod\n",
    "from dimod.generators.constraints import combinations\n",
    "from dimod import BinaryQuadraticModel\n",
    "\n",
    "from hybrid.reference import KerberosSampler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "b33ea595",
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "def solve(qubo_matrix):\n",
    "    \"\"\"Solve BQM and return matrix with solution.\"\"\"\n",
    "    bqm = dimod.BinaryQuadraticModel(qubo_matrix, \"BINARY\")\n",
    "    solution = KerberosSampler().sample(bqm,\n",
    "                                        max_iter=10,\n",
    "                                        convergence=3,\n",
    "                                        qpu_params={'label': 'Trial - Workflow Optimization'})\n",
    "    best_solution = solution.first.sample\n",
    "    solution_list = [k for k, v in best_solution.items() if v == 1]\n",
    "\n",
    "    return solution\n",
    "\n",
    "    # result = solve_sudoku(bqm, matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "7c78a0a9",
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "# solution = solve(QUBO)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "70553e86",
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "# solution"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "44b04368",
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "source": [
    "# Trying CQM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "98449079",
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['__class__',\n",
      " '__delattr__',\n",
      " '__dict__',\n",
      " '__dir__',\n",
      " '__doc__',\n",
      " '__eq__',\n",
      " '__format__',\n",
      " '__ge__',\n",
      " '__getattribute__',\n",
      " '__gt__',\n",
      " '__hash__',\n",
      " '__init__',\n",
      " '__init_subclass__',\n",
      " '__le__',\n",
      " '__lt__',\n",
      " '__module__',\n",
      " '__ne__',\n",
      " '__new__',\n",
      " '__reduce__',\n",
      " '__reduce_ex__',\n",
      " '__repr__',\n",
      " '__setattr__',\n",
      " '__sizeof__',\n",
      " '__str__',\n",
      " '__subclasshook__',\n",
      " '__weakref__',\n",
      " '_add_variables_from',\n",
      " '_discrete',\n",
      " '_iterable_to_qm',\n",
      " '_objective',\n",
      " '_substitute_self_loops_from_model',\n",
      " 'add_constraint',\n",
      " 'add_constraint_from_comparison',\n",
      " 'add_constraint_from_iterable',\n",
      " 'add_constraint_from_model',\n",
      " 'add_discrete',\n",
      " 'add_variable',\n",
      " 'check_feasible',\n",
      " 'constraints',\n",
      " 'discrete',\n",
      " 'fix_variable',\n",
      " 'fix_variables',\n",
      " 'from_bqm',\n",
      " 'from_discrete_quadratic_model',\n",
      " 'from_dqm',\n",
      " 'from_file',\n",
      " 'from_lp_file',\n",
      " 'from_qm',\n",
      " 'from_quadratic_model',\n",
      " 'is_almost_equal',\n",
      " 'is_equal',\n",
      " 'iter_constraint_data',\n",
      " 'iter_violations',\n",
      " 'lower_bound',\n",
      " 'num_biases',\n",
      " 'num_quadratic_variables',\n",
      " 'objective',\n",
      " 'relabel_variables',\n",
      " 'set_objective',\n",
      " 'substitute_self_loops',\n",
      " 'to_file',\n",
      " 'upper_bound',\n",
      " 'variables',\n",
      " 'vartype',\n",
      " 'violations']\n"
     ]
    }
   ],
   "source": [
    "from dimod import Binary, Integer, ConstrainedQuadraticModel\n",
    "from dwave.system import LeapHybridBQMSampler, LeapHybridCQMSampler\n",
    "\n",
    "x = Binary('x')\n",
    "i = Integer('i')\n",
    "j = Integer('j')\n",
    "cqm = ConstrainedQuadraticModel()\n",
    "\n",
    "pprint(dir(cqm))\n",
    "\n",
    "cqm.set_objective(2*i + 0.5*i*j -j*x + 10)\n",
    "cqm.add_constraint(x*j <= 3)                   \n",
    "cqm.add_constraint(i + j >= 1)\n",
    "\n",
    "sampler_cqm = LeapHybridCQMSampler()\n",
    "\n",
    "solution = sampler_cqm.sample_cqm(cqm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "5680273b",
   "metadata": {},
   "outputs": [],
   "source": [
    "best_solution = solution.first.sample"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "48510c3d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'i': 20046.0, 'j': 1.0, 'x': 1.0}"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_solution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "5b389b8b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['_REQUIRED_FIELDS',\n",
       " '__abstractmethods__',\n",
       " '__class__',\n",
       " '__delattr__',\n",
       " '__dict__',\n",
       " '__dir__',\n",
       " '__doc__',\n",
       " '__eq__',\n",
       " '__format__',\n",
       " '__ge__',\n",
       " '__getattribute__',\n",
       " '__getstate__',\n",
       " '__gt__',\n",
       " '__hash__',\n",
       " '__init__',\n",
       " '__init_subclass__',\n",
       " '__iter__',\n",
       " '__le__',\n",
       " '__len__',\n",
       " '__lt__',\n",
       " '__module__',\n",
       " '__ne__',\n",
       " '__new__',\n",
       " '__reduce__',\n",
       " '__reduce_ex__',\n",
       " '__repr__',\n",
       " '__setattr__',\n",
       " '__sizeof__',\n",
       " '__slots__',\n",
       " '__str__',\n",
       " '__subclasshook__',\n",
       " '__weakref__',\n",
       " '_abc_impl',\n",
       " '_asdict',\n",
       " '_info',\n",
       " '_record',\n",
       " '_variables',\n",
       " '_vartype',\n",
       " 'aggregate',\n",
       " 'append_variables',\n",
       " 'change_vartype',\n",
       " 'copy',\n",
       " 'data',\n",
       " 'data_vectors',\n",
       " 'done',\n",
       " 'filter',\n",
       " 'first',\n",
       " 'from_future',\n",
       " 'from_samples',\n",
       " 'from_samples_bqm',\n",
       " 'from_serializable',\n",
       " 'info',\n",
       " 'is_writeable',\n",
       " 'lowest',\n",
       " 'record',\n",
       " 'relabel_variables',\n",
       " 'resolve',\n",
       " 'samples',\n",
       " 'slice',\n",
       " 'to_pandas_dataframe',\n",
       " 'to_serializable',\n",
       " 'truncate',\n",
       " 'variables',\n",
       " 'vartype',\n",
       " 'wait_id']"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dir(solution)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bb5c119e",
   "metadata": {},
   "source": [
    "# Trying worflow problem using CQM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "ef8da95b",
   "metadata": {},
   "outputs": [],
   "source": [
    "cqm = ConstrainedQuadraticModel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "ad78adb6",
   "metadata": {},
   "outputs": [],
   "source": [
    "job_count = len(cost_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "a880577d",
   "metadata": {},
   "outputs": [],
   "source": [
    "xs = []\n",
    "for machines in range(len(cost_df.columns)):\n",
    "    x = [Binary(f'm{machines}_x{i}') for i in range(job_count)]\n",
    "    xs.extend(x)\n",
    "\n",
    "model = None\n",
    "for variable, cost in zip(np.array(cost_df).flatten(), xs):\n",
    "    if model:\n",
    "        model = model + cost * variable\n",
    "    else:\n",
    "        model = cost * variable\n",
    "cqm.set_objective(model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "394fb6ec",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "QuadraticModel({'m0_x0': 335.8274782608695, 'm0_x1': 159.9599304347826, 'm0_x2': 188.35060869565214, 'm0_x3': 99.97495652173913, 'm0_x4': 117.71913043478261, 'm0_x5': 159221.1884521739, 'm0_x6': 78487.56076521738, 'm0_x7': 96471.69382608694, 'm0_x8': 49054.72547826087, 'm0_x9': 60294.80864130434, 'm0_x10': 873.7241043478261, 'm0_x11': 369.07311304347826, 'm0_x12': 362.48252173913033, 'm0_x13': 230.67069565217392, 'm0_x14': 226.5515760869565, 'm0_x15': 46585.664000000004, 'm0_x16': 22726.249600000003, 'm0_x17': 27581.546000000006, 'm0_x18': 14203.906, 'm0_x19': 17238.466250000005, 'm0_x20': 782.0377043478261, 'm0_x21': 330.34351304347825, 'm1_x0': 324.4445217391304, 'm1_x1': 206.4646956521739, 'm1_x2': 202.7778260869565, 'm1_x3': 87037.90747826088, 'm1_x4': 42741.82956521739, 'm1_x5': 52293.98086956522, 'm1_x6': 26713.64347826087, 'm1_x7': 32683.73804347826, 'm1_x8': 860.0932173913044, 'm1_x9': 358.37217391304347, 'm1_x10': 343.44, 'm1_x11': 223.98260869565217, 'm1_x12': 214.65, 'm1_x13': 39946.6468173913, 'm1_x14': 19447.70963478261, 'm1_x15': 23543.116956521735, 'm1_x16': 12154.818521739131, 'm1_x17': 14714.448097826087, 'm1_x18': 673.5293217391305, 'm1_x19': 280.6372173913044, 'm1_x20': 268.944, 'm1_x21': 175.39826086956523, 'm2_x0': 168.09000000000003, 'm2_x1': 114676.08709565215, 'm2_x2': 56526.65613913043, 'm2_x3': 69475.04686956521, 'm2_x4': 35329.160086956515, 'm2_x5': 43421.904293478256, 'm2_x6': 651.3357913043478, 'm2_x7': 271.38991304347826, 'm2_x8': 260.082, 'm2_x9': 169.61869565217387, 'm2_x10': 162.55124999999998, 'm2_x11': 68406.29314782609, 'm2_x12': 33522.110469565225, 'm2_x13': 40909.46043478261, 'm2_x14': 20951.319043478263, 'm2_x15': 25568.41277173913, 'm2_x16': 714.6707478260869, 'm2_x17': 297.7794782608695, 'm2_x18': 285.37199999999996, 'm2_x19': 186.11217391304345, 'm2_x20': 178.3575, 'm2_x21': 35153.92111304348, 'm3_x0': 17079.49940869565, 'm3_x1': 20623.91008695652, 'm3_x2': 10674.68713043478, 'm3_x3': 12889.943804347826, 'm3_x4': 479.37725217391295, 'm3_x5': 199.74052173913037, 'm3_x6': 191.41799999999995, 'm3_x7': 124.83782608695648, 'm3_x8': 119.63624999999998, 'm3_x9': 149033.9595130435, 'm3_x10': 73465.79979130437, 'm3_x11': 90299.27895652175, 'm3_x12': 45916.12486956522, 'm3_x13': 56437.04934782609, 'm3_x14': 1016.6349913043479, 'm3_x15': 423.59791304347823, 'm3_x16': 405.948, 'm3_x17': 264.7486956521739, 'm3_x18': 253.71750000000003, 'm3_x19': 117514.3424, 'm3_x20': 57872.86594782608, 'm3_x21': 71051.47060869564, 'm4_x0': 36170.5412173913, 'm4_x1': 44407.16913043478, 'm4_x2': 974.7572869565217, 'm4_x3': 406.14886956521735, 'm4_x4': 389.226, 'm4_x5': 253.84304347826085, 'm4_x6': 243.26624999999999, 'm4_x7': 164060.82782608696, 'm4_x8': 80876.8612173913, 'm4_x9': 99413.81217391304, 'm4_x10': 50548.03826086956, 'm4_x11': 62133.632608695654, 'm4_x12': 1363.8049391304348, 'm4_x13': 576.0900173913044, 'm4_x14': 565.8026956521738, 'm4_x15': 360.05626086956516, 'm4_x16': 353.62668478260866, 'm4_x17': 0.384, 'm4_x18': 0.048, 'm4_x19': -0.15, 'm4_x20': 0.03, 'm4_x21': -0.09375}, {}, 0.0, {'m0_x0': 'BINARY', 'm0_x1': 'BINARY', 'm0_x2': 'BINARY', 'm0_x3': 'BINARY', 'm0_x4': 'BINARY', 'm0_x5': 'BINARY', 'm0_x6': 'BINARY', 'm0_x7': 'BINARY', 'm0_x8': 'BINARY', 'm0_x9': 'BINARY', 'm0_x10': 'BINARY', 'm0_x11': 'BINARY', 'm0_x12': 'BINARY', 'm0_x13': 'BINARY', 'm0_x14': 'BINARY', 'm0_x15': 'BINARY', 'm0_x16': 'BINARY', 'm0_x17': 'BINARY', 'm0_x18': 'BINARY', 'm0_x19': 'BINARY', 'm0_x20': 'BINARY', 'm0_x21': 'BINARY', 'm1_x0': 'BINARY', 'm1_x1': 'BINARY', 'm1_x2': 'BINARY', 'm1_x3': 'BINARY', 'm1_x4': 'BINARY', 'm1_x5': 'BINARY', 'm1_x6': 'BINARY', 'm1_x7': 'BINARY', 'm1_x8': 'BINARY', 'm1_x9': 'BINARY', 'm1_x10': 'BINARY', 'm1_x11': 'BINARY', 'm1_x12': 'BINARY', 'm1_x13': 'BINARY', 'm1_x14': 'BINARY', 'm1_x15': 'BINARY', 'm1_x16': 'BINARY', 'm1_x17': 'BINARY', 'm1_x18': 'BINARY', 'm1_x19': 'BINARY', 'm1_x20': 'BINARY', 'm1_x21': 'BINARY', 'm2_x0': 'BINARY', 'm2_x1': 'BINARY', 'm2_x2': 'BINARY', 'm2_x3': 'BINARY', 'm2_x4': 'BINARY', 'm2_x5': 'BINARY', 'm2_x6': 'BINARY', 'm2_x7': 'BINARY', 'm2_x8': 'BINARY', 'm2_x9': 'BINARY', 'm2_x10': 'BINARY', 'm2_x11': 'BINARY', 'm2_x12': 'BINARY', 'm2_x13': 'BINARY', 'm2_x14': 'BINARY', 'm2_x15': 'BINARY', 'm2_x16': 'BINARY', 'm2_x17': 'BINARY', 'm2_x18': 'BINARY', 'm2_x19': 'BINARY', 'm2_x20': 'BINARY', 'm2_x21': 'BINARY', 'm3_x0': 'BINARY', 'm3_x1': 'BINARY', 'm3_x2': 'BINARY', 'm3_x3': 'BINARY', 'm3_x4': 'BINARY', 'm3_x5': 'BINARY', 'm3_x6': 'BINARY', 'm3_x7': 'BINARY', 'm3_x8': 'BINARY', 'm3_x9': 'BINARY', 'm3_x10': 'BINARY', 'm3_x11': 'BINARY', 'm3_x12': 'BINARY', 'm3_x13': 'BINARY', 'm3_x14': 'BINARY', 'm3_x15': 'BINARY', 'm3_x16': 'BINARY', 'm3_x17': 'BINARY', 'm3_x18': 'BINARY', 'm3_x19': 'BINARY', 'm3_x20': 'BINARY', 'm3_x21': 'BINARY', 'm4_x0': 'BINARY', 'm4_x1': 'BINARY', 'm4_x2': 'BINARY', 'm4_x3': 'BINARY', 'm4_x4': 'BINARY', 'm4_x5': 'BINARY', 'm4_x6': 'BINARY', 'm4_x7': 'BINARY', 'm4_x8': 'BINARY', 'm4_x9': 'BINARY', 'm4_x10': 'BINARY', 'm4_x11': 'BINARY', 'm4_x12': 'BINARY', 'm4_x13': 'BINARY', 'm4_x14': 'BINARY', 'm4_x15': 'BINARY', 'm4_x16': 'BINARY', 'm4_x17': 'BINARY', 'm4_x18': 'BINARY', 'm4_x19': 'BINARY', 'm4_x20': 'BINARY', 'm4_x21': 'BINARY'}, dtype='float64')"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cqm.objective"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1da52047",
   "metadata": {},
   "source": [
    "### must use one machine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "f8d4249f",
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(job_count):\n",
    "    one_machine = None\n",
    "    for j in range(i, len(xs), job_count):\n",
    "        if one_machine:\n",
    "            one_machine = one_machine + xs[j]\n",
    "        else:\n",
    "            one_machine = xs[j]\n",
    "    cqm.add_constraint( one_machine == 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b074f305",
   "metadata": {},
   "source": [
    "### All paths must finish before a deadline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "98ecb242",
   "metadata": {},
   "outputs": [],
   "source": [
    "flat_runtimes = [(runtime, name) for n, machine_runtimes in runtimes.items() for runtime, name in zip(machine_runtimes, [j.name for j in jobs])]\n",
    "\n",
    "for path in paths:\n",
    "    path_runtime = None\n",
    "    for var, (runtime, name) in zip(xs, flat_runtimes):\n",
    "        if name not in path:\n",
    "            continue\n",
    "        if path_runtime:\n",
    "            path_runtime = path_runtime + runtime * var\n",
    "        else:\n",
    "            path_runtime = runtime * var\n",
    "    cqm.add_constraint(path_runtime <= DEADLINE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "d8c6ccc6",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'5bc5a6': BinaryQuadraticModel({'m0_x0': 1.0, 'm1_x0': 1.0, 'm2_x0': 1.0, 'm3_x0': 1.0, 'm4_x0': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '5ca37e': BinaryQuadraticModel({'m0_x1': 1.0, 'm1_x1': 1.0, 'm2_x1': 1.0, 'm3_x1': 1.0, 'm4_x1': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '7fa3e2': BinaryQuadraticModel({'m0_x2': 1.0, 'm1_x2': 1.0, 'm2_x2': 1.0, 'm3_x2': 1.0, 'm4_x2': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " 'aadb18': BinaryQuadraticModel({'m0_x3': 1.0, 'm1_x3': 1.0, 'm2_x3': 1.0, 'm3_x3': 1.0, 'm4_x3': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '00e7f8': BinaryQuadraticModel({'m0_x4': 1.0, 'm1_x4': 1.0, 'm2_x4': 1.0, 'm3_x4': 1.0, 'm4_x4': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '74153f': BinaryQuadraticModel({'m0_x5': 1.0, 'm1_x5': 1.0, 'm2_x5': 1.0, 'm3_x5': 1.0, 'm4_x5': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " 'cc45d5': BinaryQuadraticModel({'m0_x6': 1.0, 'm1_x6': 1.0, 'm2_x6': 1.0, 'm3_x6': 1.0, 'm4_x6': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " 'd5ccdc': BinaryQuadraticModel({'m0_x7': 1.0, 'm1_x7': 1.0, 'm2_x7': 1.0, 'm3_x7': 1.0, 'm4_x7': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '12de41': BinaryQuadraticModel({'m0_x8': 1.0, 'm1_x8': 1.0, 'm2_x8': 1.0, 'm3_x8': 1.0, 'm4_x8': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '5b16ee': BinaryQuadraticModel({'m0_x9': 1.0, 'm1_x9': 1.0, 'm2_x9': 1.0, 'm3_x9': 1.0, 'm4_x9': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " 'be0c39': BinaryQuadraticModel({'m0_x10': 1.0, 'm1_x10': 1.0, 'm2_x10': 1.0, 'm3_x10': 1.0, 'm4_x10': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " 'da0798': BinaryQuadraticModel({'m0_x11': 1.0, 'm1_x11': 1.0, 'm2_x11': 1.0, 'm3_x11': 1.0, 'm4_x11': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '93e9b4': BinaryQuadraticModel({'m0_x12': 1.0, 'm1_x12': 1.0, 'm2_x12': 1.0, 'm3_x12': 1.0, 'm4_x12': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '5f2672': BinaryQuadraticModel({'m0_x13': 1.0, 'm1_x13': 1.0, 'm2_x13': 1.0, 'm3_x13': 1.0, 'm4_x13': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '88f178': BinaryQuadraticModel({'m0_x14': 1.0, 'm1_x14': 1.0, 'm2_x14': 1.0, 'm3_x14': 1.0, 'm4_x14': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '608238': BinaryQuadraticModel({'m0_x15': 1.0, 'm1_x15': 1.0, 'm2_x15': 1.0, 'm3_x15': 1.0, 'm4_x15': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '6db733': BinaryQuadraticModel({'m0_x16': 1.0, 'm1_x16': 1.0, 'm2_x16': 1.0, 'm3_x16': 1.0, 'm4_x16': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '48b2ed': BinaryQuadraticModel({'m0_x17': 1.0, 'm1_x17': 1.0, 'm2_x17': 1.0, 'm3_x17': 1.0, 'm4_x17': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '00a752': BinaryQuadraticModel({'m0_x18': 1.0, 'm1_x18': 1.0, 'm2_x18': 1.0, 'm3_x18': 1.0, 'm4_x18': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '0ead40': BinaryQuadraticModel({'m0_x19': 1.0, 'm1_x19': 1.0, 'm2_x19': 1.0, 'm3_x19': 1.0, 'm4_x19': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '6a9865': BinaryQuadraticModel({'m0_x20': 1.0, 'm1_x20': 1.0, 'm2_x20': 1.0, 'm3_x20': 1.0, 'm4_x20': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " 'a3994f': BinaryQuadraticModel({'m0_x21': 1.0, 'm1_x21': 1.0, 'm2_x21': 1.0, 'm3_x21': 1.0, 'm4_x21': 1.0}, {}, 0.0, 'BINARY') == 1,\n",
       " '3734a9': BinaryQuadraticModel({'m0_x0': 220.93913043478258, 'm0_x2': 1883.0260869565216, 'm0_x21': 4.0, 'm1_x0': 55.234782608695646, 'm1_x2': 470.7565217391304, 'm1_x21': 1.0, 'm2_x0': 34.521739130434774, 'm2_x2': 294.2228260869565, 'm2_x21': 0.625, 'm3_x0': 0.5523478260869564, 'm3_x2': 4.707565217391304, 'm3_x21': 0.01, 'm4_x0': 0.3452173913043478, 'm4_x2': 2.942228260869565, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " 'ac94a7': BinaryQuadraticModel({'m0_x0': 220.93913043478258, 'm0_x4': 1685.4260869565217, 'm0_x21': 4.0, 'm1_x0': 55.234782608695646, 'm1_x4': 421.3565217391304, 'm1_x21': 1.0, 'm2_x0': 34.521739130434774, 'm2_x4': 263.3478260869565, 'm2_x21': 0.625, 'm3_x0': 0.5523478260869564, 'm3_x4': 4.213565217391304, 'm3_x21': 0.01, 'm4_x0': 0.3452173913043478, 'm4_x4': 2.6334782608695653, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " '3e63b1': BinaryQuadraticModel({'m0_x0': 220.93913043478258, 'm0_x6': 1990.9565217391305, 'm0_x21': 4.0, 'm1_x0': 55.234782608695646, 'm1_x6': 497.7391304347826, 'm1_x21': 1.0, 'm2_x0': 34.521739130434774, 'm2_x6': 311.0869565217391, 'm2_x21': 0.625, 'm3_x0': 0.5523478260869564, 'm3_x6': 4.977391304347826, 'm3_x21': 0.01, 'm4_x0': 0.3452173913043478, 'm4_x6': 3.110869565217391, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " '26b66a': BinaryQuadraticModel({'m0_x0': 220.93913043478258, 'm0_x8': 1559.095652173913, 'm0_x21': 4.0, 'm1_x0': 55.234782608695646, 'm1_x8': 389.7739130434783, 'm1_x21': 1.0, 'm2_x0': 34.521739130434774, 'm2_x8': 243.6086956521739, 'm2_x21': 0.625, 'm3_x0': 0.5523478260869564, 'm3_x8': 3.8977391304347826, 'm3_x21': 0.01, 'm4_x0': 0.3452173913043478, 'm4_x8': 2.436086956521739, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " '12dd09': BinaryQuadraticModel({'m0_x0': 220.93913043478258, 'm0_x10': 1507.7217391304346, 'm0_x21': 4.0, 'm1_x0': 55.234782608695646, 'm1_x10': 376.93043478260864, 'm1_x21': 1.0, 'm2_x0': 34.521739130434774, 'm2_x10': 235.58152173913038, 'm2_x21': 0.625, 'm3_x0': 0.5523478260869564, 'm3_x10': 3.7693043478260866, 'm3_x21': 0.01, 'm4_x0': 0.3452173913043478, 'm4_x10': 2.355815217391304, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " '9c3a33': BinaryQuadraticModel({'m0_x0': 220.93913043478258, 'm0_x12': 1654.3304347826083, 'm0_x21': 4.0, 'm1_x0': 55.234782608695646, 'm1_x12': 413.5826086956521, 'm1_x21': 1.0, 'm2_x0': 34.521739130434774, 'm2_x12': 258.4891304347825, 'm2_x21': 0.625, 'm3_x0': 0.5523478260869564, 'm3_x12': 4.135826086956521, 'm3_x21': 0.01, 'm4_x0': 0.3452173913043478, 'm4_x12': 2.5848913043478254, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " 'de5c5b': BinaryQuadraticModel({'m0_x0': 220.93913043478258, 'm0_x14': 1109.669565217391, 'm0_x21': 4.0, 'm1_x0': 55.234782608695646, 'm1_x14': 277.41739130434775, 'm1_x21': 1.0, 'm2_x0': 34.521739130434774, 'm2_x14': 173.38586956521732, 'm2_x21': 0.625, 'm3_x0': 0.5523478260869564, 'm3_x14': 2.7741739130434775, 'm3_x21': 0.01, 'm4_x0': 0.3452173913043478, 'm4_x14': 1.7338586956521733, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " '47fa73': BinaryQuadraticModel({'m0_x0': 220.93913043478258, 'm0_x16': 2353.3217391304347, 'm0_x21': 4.0, 'm1_x0': 55.234782608695646, 'm1_x16': 588.3304347826087, 'm1_x21': 1.0, 'm2_x0': 34.521739130434774, 'm2_x16': 367.7065217391304, 'm2_x21': 0.625, 'm3_x0': 0.5523478260869564, 'm3_x16': 5.883304347826087, 'm3_x21': 0.01, 'm4_x0': 0.3452173913043478, 'm4_x16': 3.6770652173913043, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " 'cc293c': BinaryQuadraticModel({'m0_x0': 220.93913043478258, 'm0_x18': 2256.382608695652, 'm0_x21': 4.0, 'm1_x0': 55.234782608695646, 'm1_x18': 564.095652173913, 'm1_x21': 1.0, 'm2_x0': 34.521739130434774, 'm2_x18': 352.55978260869557, 'm2_x21': 0.625, 'm3_x0': 0.5523478260869564, 'm3_x18': 5.64095652173913, 'm3_x21': 0.01, 'm4_x0': 0.3452173913043478, 'm4_x18': 3.525597826086956, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " '689131': BinaryQuadraticModel({'m0_x0': 220.93913043478258, 'm0_x20': 2939.2347826086952, 'm0_x21': 4.0, 'm1_x0': 55.234782608695646, 'm1_x20': 734.8086956521738, 'm1_x21': 1.0, 'm2_x0': 34.521739130434774, 'm2_x20': 459.25543478260863, 'm2_x21': 0.625, 'm3_x0': 0.5523478260869564, 'm3_x20': 7.348086956521738, 'm3_x21': 0.01, 'm4_x0': 0.3452173913043478, 'm4_x20': 4.592554347826086, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " '4eef9e': BinaryQuadraticModel({'m0_x1': 31195.373913043473, 'm0_x2': 1883.0260869565216, 'm0_x21': 4.0, 'm1_x1': 7798.843478260868, 'm1_x2': 470.7565217391304, 'm1_x21': 1.0, 'm2_x1': 4874.277173913042, 'm2_x2': 294.2228260869565, 'm2_x21': 0.625, 'm3_x1': 77.98843478260868, 'm3_x2': 4.707565217391304, 'm3_x21': 0.01, 'm4_x1': 48.742771739130426, 'm4_x2': 2.942228260869565, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " '31bd20': BinaryQuadraticModel({'m0_x3': 15738.400000000001, 'm0_x4': 1685.4260869565217, 'm0_x21': 4.0, 'm1_x3': 3934.6000000000004, 'm1_x4': 421.3565217391304, 'm1_x21': 1.0, 'm2_x3': 2459.125, 'm2_x4': 263.3478260869565, 'm2_x21': 0.625, 'm3_x3': 39.346000000000004, 'm3_x4': 4.213565217391304, 'm3_x21': 0.01, 'm4_x3': 24.591250000000002, 'm4_x4': 2.6334782608695653, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " '73d493': BinaryQuadraticModel({'m0_x5': 21586.782608695652, 'm0_x6': 1990.9565217391305, 'm0_x21': 4.0, 'm1_x5': 5396.695652173913, 'm1_x6': 497.7391304347826, 'm1_x21': 1.0, 'm2_x5': 3372.9347826086955, 'm2_x6': 311.0869565217391, 'm2_x21': 0.625, 'm3_x5': 53.96695652173913, 'm3_x6': 4.977391304347826, 'm3_x21': 0.01, 'm4_x5': 33.72934782608696, 'm4_x6': 3.110869565217391, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " '1f2efe': BinaryQuadraticModel({'m0_x7': 14600.38260869565, 'm0_x8': 1559.095652173913, 'm0_x21': 4.0, 'm1_x7': 3650.0956521739126, 'm1_x8': 389.7739130434783, 'm1_x21': 1.0, 'm2_x7': 2281.309782608695, 'm2_x8': 243.6086956521739, 'm2_x21': 0.625, 'm3_x7': 36.50095652173913, 'm3_x8': 3.8977391304347826, 'm3_x21': 0.01, 'm4_x7': 22.813097826086953, 'm4_x8': 2.436086956521739, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " 'd223a6': BinaryQuadraticModel({'m0_x9': 22538.53913043478, 'm0_x10': 1507.7217391304346, 'm0_x21': 4.0, 'm1_x9': 5634.634782608695, 'm1_x10': 376.93043478260864, 'm1_x21': 1.0, 'm2_x9': 3521.646739130434, 'm2_x10': 235.58152173913038, 'm2_x21': 0.625, 'm3_x9': 56.34634782608695, 'm3_x10': 3.7693043478260866, 'm3_x21': 0.01, 'm4_x9': 35.21646739130434, 'm4_x10': 2.355815217391304, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " 'd64c36': BinaryQuadraticModel({'m0_x11': 18917.66956521739, 'm0_x12': 1654.3304347826083, 'm0_x21': 4.0, 'm1_x11': 4729.417391304348, 'm1_x12': 413.5826086956521, 'm1_x21': 1.0, 'm2_x11': 2955.885869565217, 'm2_x12': 258.4891304347825, 'm2_x21': 0.625, 'm3_x11': 47.29417391304348, 'm3_x12': 4.135826086956521, 'm3_x21': 0.01, 'm4_x11': 29.558858695652173, 'm4_x12': 2.5848913043478254, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " 'd25b4e': BinaryQuadraticModel({'m0_x13': 13818.365217391303, 'm0_x14': 1109.669565217391, 'm0_x21': 4.0, 'm1_x13': 3454.591304347826, 'm1_x14': 277.41739130434775, 'm1_x21': 1.0, 'm2_x13': 2159.119565217391, 'm2_x14': 173.38586956521732, 'm2_x21': 0.625, 'm3_x13': 34.54591304347826, 'm3_x14': 2.7741739130434775, 'm3_x21': 0.01, 'm4_x13': 21.591195652173912, 'm4_x14': 1.7338586956521733, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " 'aafc32': BinaryQuadraticModel({'m0_x15': 29199.44347826087, 'm0_x16': 2353.3217391304347, 'm0_x21': 4.0, 'm1_x15': 7299.860869565217, 'm1_x16': 588.3304347826087, 'm1_x21': 1.0, 'm2_x15': 4562.413043478261, 'm2_x16': 367.7065217391304, 'm2_x21': 0.625, 'm3_x15': 72.99860869565218, 'm3_x16': 5.883304347826087, 'm3_x21': 0.01, 'm4_x15': 45.62413043478261, 'm4_x16': 3.6770652173913043, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " '8052a2': BinaryQuadraticModel({'m0_x17': 24564.03478260869, 'm0_x18': 2256.382608695652, 'm0_x21': 4.0, 'm1_x17': 6141.008695652173, 'm1_x18': 564.095652173913, 'm1_x21': 1.0, 'm2_x17': 3838.1304347826076, 'm2_x18': 352.55978260869557, 'm2_x21': 0.625, 'm3_x17': 61.410086956521724, 'm3_x18': 5.64095652173913, 'm3_x21': 0.01, 'm4_x17': 38.38130434782608, 'm4_x18': 3.525597826086956, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000,\n",
       " '3ebcab': BinaryQuadraticModel({'m0_x19': 32043.130434782608, 'm0_x20': 2939.2347826086952, 'm0_x21': 4.0, 'm1_x19': 8010.782608695652, 'm1_x20': 734.8086956521738, 'm1_x21': 1.0, 'm2_x19': 5006.739130434782, 'm2_x20': 459.25543478260863, 'm2_x21': 0.625, 'm3_x19': 80.10782608695652, 'm3_x20': 7.348086956521738, 'm3_x21': 0.01, 'm4_x19': 50.06739130434782, 'm4_x20': 4.592554347826086, 'm4_x21': 0.00625}, {}, 0.0, 'BINARY') <= 40000}"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cqm.constraints"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "cb147ec6",
   "metadata": {},
   "outputs": [],
   "source": [
    "sampler_cqm = LeapHybridCQMSampler()\n",
    "\n",
    "solution = sampler_cqm.sample_cqm(cqm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "47fff72c",
   "metadata": {},
   "outputs": [],
   "source": [
    "best_solution = solution.first.sample"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "e4bc625c",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'m0_x0': 0.0,\n",
       " 'm0_x1': 0.0,\n",
       " 'm0_x10': 0.0,\n",
       " 'm0_x11': 0.0,\n",
       " 'm0_x12': 0.0,\n",
       " 'm0_x13': 1.0,\n",
       " 'm0_x14': 0.0,\n",
       " 'm0_x15': 0.0,\n",
       " 'm0_x16': 0.0,\n",
       " 'm0_x17': 0.0,\n",
       " 'm0_x18': 0.0,\n",
       " 'm0_x19': 0.0,\n",
       " 'm0_x2': 1.0,\n",
       " 'm0_x20': 0.0,\n",
       " 'm0_x21': 0.0,\n",
       " 'm0_x3': 1.0,\n",
       " 'm0_x4': 1.0,\n",
       " 'm0_x5': 0.0,\n",
       " 'm0_x6': 0.0,\n",
       " 'm0_x7': 0.0,\n",
       " 'm0_x8': 0.0,\n",
       " 'm0_x9': 0.0,\n",
       " 'm1_x0': 0.0,\n",
       " 'm1_x1': 0.0,\n",
       " 'm1_x10': 0.0,\n",
       " 'm1_x11': 1.0,\n",
       " 'm1_x12': 1.0,\n",
       " 'm1_x13': 0.0,\n",
       " 'm1_x14': 0.0,\n",
       " 'm1_x15': 0.0,\n",
       " 'm1_x16': 0.0,\n",
       " 'm1_x17': 0.0,\n",
       " 'm1_x18': 0.0,\n",
       " 'm1_x19': 0.0,\n",
       " 'm1_x2': 0.0,\n",
       " 'm1_x20': 0.0,\n",
       " 'm1_x21': 0.0,\n",
       " 'm1_x3': 0.0,\n",
       " 'm1_x4': 0.0,\n",
       " 'm1_x5': 0.0,\n",
       " 'm1_x6': 0.0,\n",
       " 'm1_x7': 0.0,\n",
       " 'm1_x8': 0.0,\n",
       " 'm1_x9': 0.0,\n",
       " 'm2_x0': 1.0,\n",
       " 'm2_x1': 0.0,\n",
       " 'm2_x10': 0.0,\n",
       " 'm2_x11': 0.0,\n",
       " 'm2_x12': 0.0,\n",
       " 'm2_x13': 0.0,\n",
       " 'm2_x14': 0.0,\n",
       " 'm2_x15': 0.0,\n",
       " 'm2_x16': 0.0,\n",
       " 'm2_x17': 0.0,\n",
       " 'm2_x18': 0.0,\n",
       " 'm2_x19': 0.0,\n",
       " 'm2_x2': 0.0,\n",
       " 'm2_x20': 0.0,\n",
       " 'm2_x21': 0.0,\n",
       " 'm2_x3': 0.0,\n",
       " 'm2_x4': 0.0,\n",
       " 'm2_x5': 0.0,\n",
       " 'm2_x6': 0.0,\n",
       " 'm2_x7': 0.0,\n",
       " 'm2_x8': 0.0,\n",
       " 'm2_x9': 1.0,\n",
       " 'm3_x0': 0.0,\n",
       " 'm3_x1': 0.0,\n",
       " 'm3_x10': 0.0,\n",
       " 'm3_x11': 0.0,\n",
       " 'm3_x12': 0.0,\n",
       " 'm3_x13': 0.0,\n",
       " 'm3_x14': 0.0,\n",
       " 'm3_x15': 0.0,\n",
       " 'm3_x16': 0.0,\n",
       " 'm3_x17': 0.0,\n",
       " 'm3_x18': 0.0,\n",
       " 'm3_x19': 0.0,\n",
       " 'm3_x2': 0.0,\n",
       " 'm3_x20': 0.0,\n",
       " 'm3_x21': 0.0,\n",
       " 'm3_x3': 0.0,\n",
       " 'm3_x4': 0.0,\n",
       " 'm3_x5': 0.0,\n",
       " 'm3_x6': 1.0,\n",
       " 'm3_x7': 1.0,\n",
       " 'm3_x8': 1.0,\n",
       " 'm3_x9': 0.0,\n",
       " 'm4_x0': 0.0,\n",
       " 'm4_x1': 0.0,\n",
       " 'm4_x10': 0.0,\n",
       " 'm4_x11': 0.0,\n",
       " 'm4_x12': 0.0,\n",
       " 'm4_x13': 0.0,\n",
       " 'm4_x14': 0.0,\n",
       " 'm4_x15': 0.0,\n",
       " 'm4_x16': 0.0,\n",
       " 'm4_x17': 1.0,\n",
       " 'm4_x18': 1.0,\n",
       " 'm4_x19': 1.0,\n",
       " 'm4_x2': 0.0,\n",
       " 'm4_x20': 1.0,\n",
       " 'm4_x21': 1.0,\n",
       " 'm4_x3': 0.0,\n",
       " 'm4_x4': 0.0,\n",
       " 'm4_x5': 0.0,\n",
       " 'm4_x6': 0.0,\n",
       " 'm4_x7': 0.0,\n",
       " 'm4_x8': 0.0,\n",
       " 'm4_x9': 0.0}"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_solution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "b4f2e73a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "provided_solutions: 52\n",
      "correct_solutions: 34\n"
     ]
    }
   ],
   "source": [
    "def is_correct_solution(cqm, sol):\n",
    "    return len(cqm.violations(sol, skip_satisfied = True)) == 0\n",
    "\n",
    "correct_solutions = [ s for s in solution if is_correct_solution(cqm, s)]\n",
    "\n",
    "print( f\"provided_solutions: {len(solution)}\")\n",
    "print( f\"correct_solutions: {len(correct_solutions)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "3eb083dc",
   "metadata": {},
   "outputs": [],
   "source": [
    "from parse import parse\n",
    "\n",
    "def solve_workflow(machines, jobs, solution):\n",
    "    actual_solution = {}\n",
    "\n",
    "    for k, is_used in solution.items():\n",
    "        machine, var = parse('m{}_x{}', k)\n",
    "        if is_used:\n",
    "            actual_solution[int(var)] = machines[int(machine)]\n",
    "    return actual_solution\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "e9745c2d",
   "metadata": {},
   "outputs": [],
   "source": [
    "machine_names = cost_df.columns\n",
    "job_names = cost_df.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "17a61095",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{1: 'ZeusCpu',\n",
       " 13: 'ZeusCpu',\n",
       " 14: 'ZeusCpu',\n",
       " 2: 'ZeusCpu',\n",
       " 3: 'ZeusCpu',\n",
       " 4: 'ZeusCpu',\n",
       " 11: 'PrometeusCpu',\n",
       " 12: 'PrometeusCpu',\n",
       " 0: 'AresCpu',\n",
       " 10: 'AresCpu',\n",
       " 9: 'AresCpu',\n",
       " 5: 'PrometeusGpu',\n",
       " 6: 'PrometeusGpu',\n",
       " 7: 'PrometeusGpu',\n",
       " 8: 'PrometeusGpu',\n",
       " 15: 'AresGpu',\n",
       " 16: 'AresGpu',\n",
       " 17: 'AresGpu',\n",
       " 18: 'AresGpu',\n",
       " 19: 'AresGpu',\n",
       " 20: 'AresGpu',\n",
       " 21: 'AresGpu'}"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "solve_workflow(machine_names, job_names, correct_solutions[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d2785536",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "33c0b469",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "Kopia notatnika [WSP] Wfcommons.ipynb",
   "provenance": []
  },
  "interpreter": {
   "hash": "31f2aee4e71d21fbe5cf8b01ff0e069b9275f58929596ceb00d14d90e3e16cd6"
  },
  "kernelspec": {
   "display_name": "Python 3.8.10 64-bit",
   "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": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}