{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "NzWFca3cB9oe"
   },
   "source": [
    "![alt text](https://easyvvuq.readthedocs.io/en/latest/_static/circle-logo.svg)\n",
    "\n",
    "Vytautas Jancauskas 2020\n",
    "\n",
    "# **EasyVVUQ Tutorial**\n",
    "\n",
    "# Introduction\n",
    "\n",
    "In this tutorial we will show you how you can use [EasyVVUQ](https://github.com/UCL-CCS/EasyVVUQ) to investigate the properties of a simple epidemiological model. The model is very simplistic and is not intended to realisticly portray a real epidemic such as COVID-19. However it is also fun enough to experiment with and we can use it as an example to show how you can use EasyVVUQ to answer questions about your scientific models. EasyVVUQ is successfully used by researchers in various different fields, such as plasma physics, weather and air pollution modelling and materials science."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ugDD5kSbAtw_"
   },
   "source": [
    "# Installing EasyVVUQ\n",
    "\n",
    "Before we do anything else we need to install EasyVVUQ for use in this notebook. Please skip this if you already have EasyVVUQ installed in your system. This is meant for situations where this notebook is hosted externally to your local computer. We also want to clone the git repository because it contains some files used in this tutorial. Please note that you might need to restart the runtime after the installation is complete. If the following code examples don't work please try that first. I have added a command that kills the runtime (which causes it to be restarted). But I'm not sure if it will always work."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:29.751332Z",
     "start_time": "2021-06-09T11:12:29.748916Z"
    }
   },
   "outputs": [],
   "source": [
    "#!pip install git+https://github.com/UCL-CCS/EasyVVUQ"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:29.755344Z",
     "start_time": "2021-06-09T11:12:29.753573Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "We-D7yY3H1ds"
   },
   "outputs": [],
   "source": [
    "#!git clone https://github.com/UCL-CCS/EasyVVUQ"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:29.758796Z",
     "start_time": "2021-06-09T11:12:29.757271Z"
    }
   },
   "outputs": [],
   "source": [
    "#import os\n",
    "#os.kill(os.getpid(), 9)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "O5RXyRXm4tIf"
   },
   "source": [
    "# Epidemiological Model\n",
    "\n",
    "In our model individuals are placed on a two dimensional square grid. During each turn, the individual moves at random one square in one of eight possible directions (N, NE, E, SE, S, SW, W or NW). If the individual encounters another individual and that person is ill, the individual will also become sick for a specified number of turns (that number is a parameter to the model). Once ill, the individual cannot become sick anymore and after getting over the disease they will have immunity. Immunity means that this person cannot get infected anymore. The simulation continues until no one is sick. At that point the disease counts as erradicated."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "chwejpvvRCNC"
   },
   "source": [
    "Let us load the model and see how it operates. We create a population on a 10x10 grid containing 20 individuals. Once infected the individual can transmit the disease for 28 turns. After that they have a 20% chance of dying."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:30.253856Z",
     "start_time": "2021-06-09T11:12:29.760385Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "cBYFOLALP7g7"
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import epidemic.epidemic as epidemic\n",
    "\n",
    "population = epidemic.Population(grid_size=10, n=20, duration=28, mortality=0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "LD0bwjvxbHbP"
   },
   "source": [
    "The code cell below is supposed to be run multiple times. Each time you run it the image below will update to show the state of the population. The black squares are empty, the white squares represent individuals who are not immune and are not sick. The red squares represent ill individuals and the intensity of the red shows how many turns they will stay ill for. Green squares represent individuals who are immune. Once the disease is eradicated, the graphic will change to a plot showing the evolution of the disease over time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:30.320154Z",
     "start_time": "2021-06-09T11:12:30.254880Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "_gDcO4TQRM-Y"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "population.move()\n",
    "if np.count_nonzero(population.ill):\n",
    "  plt.imshow(population.get_im())\n",
    "  plt.show()\n",
    "else:\n",
    "  plt.plot(population.ill_history, label=\"Sick Individuals\")\n",
    "  plt.plot(population.immune_history, label=\"Immune Individuals\")\n",
    "  plt.plot(population.n_history, label=\"Population\")\n",
    "  plt.xlabel(\"Time\")\n",
    "  plt.ylabel(\"Count\")\n",
    "  plt.legend()\n",
    "  plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "lMaeqU5942PB"
   },
   "source": [
    "Since EasyVVUQ is meant to be a general framework (non-Python specific) we don't call Python functions directly to get results of the simulation. After all, many simulations are still written in Fortran and operate by taking an input file and producing a file with outputs of the simulation. To do statistical analysis we need to be able to provide an appropriate input file to the simulation and be able to parse the outputs of the simulation. You can run our simulation as in the following example."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:31.959342Z",
     "start_time": "2021-06-09T11:12:30.321059Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "GStHHTzNAndx"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.887u 0.711s 0:01.25 207.2%\t0+0k 0+0io 525pf+0w\r\n"
     ]
    }
   ],
   "source": [
    "!cd epidemic/; time python3 epidemic.py example.json output.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "zl3ItaR5NdxV"
   },
   "source": [
    "This will have produced a file called output.csv that consists of four columns labeled \"iteration\", \"ill\", \"immune\" and \"population\". These should be fairly self explanatory."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:32.325452Z",
     "start_time": "2021-06-09T11:12:31.960859Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "UeKJH_1sBGsN",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration,ill,immune,population\r\n",
      "0,2,1,20\r\n",
      "1,2,1,20\r\n",
      "2,2,1,20\r\n",
      "3,2,1,20\r\n",
      "4,2,1,20\r\n",
      "5,2,1,20\r\n",
      "6,3,2,20\r\n",
      "7,3,2,20\r\n",
      "8,3,2,20\r\n",
      "9,3,2,20\r\n",
      "10,3,2,20\r\n",
      "11,3,2,20\r\n",
      "12,3,2,20\r\n",
      "13,4,3,20\r\n",
      "14,5,4,20\r\n",
      "15,6,5,20\r\n",
      "16,6,5,20\r\n",
      "17,7,6,20\r\n",
      "18,7,6,20\r\n",
      "19,7,6,20\r\n",
      "20,8,7,20\r\n",
      "21,9,8,20\r\n",
      "22,10,9,20\r\n",
      "23,11,10,20\r\n",
      "24,13,12,20\r\n",
      "25,14,13,20\r\n",
      "26,15,14,20\r\n",
      "27,14,15,20\r\n",
      "28,14,15,20\r\n",
      "29,14,15,20\r\n",
      "30,16,17,20\r\n",
      "31,16,17,20\r\n",
      "32,16,17,20\r\n",
      "33,16,18,20\r\n",
      "34,16,18,20\r\n",
      "35,16,18,20\r\n",
      "36,16,18,20\r\n",
      "37,16,18,20\r\n",
      "38,17,19,20\r\n",
      "39,17,19,20\r\n",
      "40,16,19,20\r\n",
      "41,16,20,20\r\n",
      "42,15,20,20\r\n",
      "43,15,20,20\r\n",
      "44,14,20,20\r\n",
      "45,14,20,20\r\n",
      "46,14,20,20\r\n",
      "47,13,20,20\r\n",
      "48,12,20,20\r\n",
      "49,11,20,20\r\n",
      "50,10,20,20\r\n",
      "51,8,20,20\r\n",
      "52,7,20,20\r\n",
      "53,6,20,20\r\n",
      "54,5,19,19\r\n",
      "55,5,19,19\r\n",
      "56,5,19,19\r\n",
      "57,3,19,19\r\n",
      "58,3,19,19\r\n",
      "59,3,19,19\r\n",
      "60,2,19,19\r\n",
      "61,2,19,19\r\n",
      "62,2,19,19\r\n",
      "63,2,19,19\r\n",
      "64,2,19,19\r\n",
      "65,1,18,18\r\n",
      "66,1,18,18\r\n",
      "67,1,18,18\r\n",
      "68,0,18,18\r\n"
     ]
    }
   ],
   "source": [
    "!cat epidemic/output.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "B-VMs_nEBBhE"
   },
   "source": [
    "Let us plot this data and see what the evolution of the disease looks like in our population."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:32.712532Z",
     "start_time": "2021-06-09T11:12:32.326927Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "Wci3iduVBr7E"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "  import matplotlib.pyplot as plt\n",
    "  import pandas as pd\n",
    "\n",
    "  df = pd.read_csv(\"epidemic/output.csv\")\n",
    "  plt.plot(df['ill'], label=\"Sick Individuals\")\n",
    "  plt.plot(df['immune'], label=\"Immune Individuals\")\n",
    "  plt.plot(df['population'], label=\"Population\")\n",
    "  plt.xlabel(\"Time\")\n",
    "  plt.ylabel(\"Count\")\n",
    "  plt.legend()\n",
    "  plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ZPTzDP7KaA4Y"
   },
   "source": [
    "A short summary is in order so that we can start exploring the sample space of our model:\n",
    "\n",
    "* We have a script that takes a JSON file with parameters and produces a CSV file with the output.\n",
    "* The model takes 4 input parameters - grid size, population size, disease duration and mortality rate.\n",
    "* The model produces 4 columns of output - iteration number, number of sick people, number of immune people and the current population size."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "1qCJk4g5bsfI"
   },
   "source": [
    "We will use EasyVVUQ to help us answer some questions about the model. Here are some simple ones that arise from toying with the model:\n",
    "\n",
    "* Given that every time the length of time before the disease is erradicated is different even with the same parameters (due to the fact that each individual chooses where to move to at random), we might want to know, within a given certainty range, what the expected value of that is. This tells us, with needed confidence, how long we can expect the disease to last given certain parameters.\n",
    "* We might also do the same thing but for a set of parameter values. Namely we might want to performa a parameter sweep with corresponding error bars.\n",
    "* We might also want to improve our results by adding more samples to our analysis - hence we will see how we can restart a simulation and draw more samples to improve the accuracy.\n",
    "* At some point we will want to use external resources to execute our simulations. We will quickly discuss how this can be done.\n",
    "* Finally, given that our model is quite computationally expensive, we might want to explore the possibility of creating surrogate models to stand-in in place of the original model. These are usually expensive to create but very cheap to evaluate. Hence there is a possibility that we will be able to extract knowledge about our model from them that would be too expensive (computationally or otherwise) with a full simulation."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "FOPcPFeJT3Vt"
   },
   "source": [
    "But first we need to set EasyVVUQ up to produce the configuration files in the suitable format and read in the output of the simulation. We also need to give a description of the parameter space. We also need to specify how we will execute our simulation. The next sections is concerened with these tasks."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "zkWCN7Pe6GJB"
   },
   "source": [
    "# EasyVVUQ Set-up\n",
    "\n",
    "For the examples in this tutorial we import some libraries that will be used throughout. EasyVVUQ will be referred to as 'uq' in the code. We also need Chaospy because we use it for the probability distribution classes. We use numpy for certain small tasks and we use pandas DataFrame as the standard data exchange format as is customary in the Python data science infrastucture."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:34.847231Z",
     "start_time": "2021-06-09T11:12:32.714371Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "Yd2XHwIE4tIg"
   },
   "outputs": [],
   "source": [
    "import easyvvuq as uq\n",
    "import chaospy as cp\n",
    "import easyvvuq.collate\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import os\n",
    "from pathlib import Path\n",
    "from easyvvuq.actions import CreateRunDirectory, Encode, Decode, CleanUp, ExecuteLocal, Actions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "PhCZgtszlpU5"
   },
   "source": [
    "While we are at it we also want to describe the arguments to our model. This takes a form of a Python dictionary. The dictionary is based on the Cerberus validator dictionary format. For more information refer to Cerberus [documentation](https://docs.python-cerberus.org/en/stable/). This dictionary is used in both validation of the results and for the default values when we don't want to vary a certain parameter."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:29:34.790630Z",
     "start_time": "2021-06-09T11:29:34.785832Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "isCmUx0bm6AX"
   },
   "outputs": [],
   "source": [
    "params = {\n",
    "    \"grid_size\": {\"type\": \"float\", \"default\": 10},\n",
    "    \"n\": {\"type\": \"float\", \"min\": 1, \"max\": 100, \"default\": 20},\n",
    "    \"duration\": {\"type\": \"float\", \"min\": 0, \"max\": 365, \"default\": 28},\n",
    "    \"mortality\": {\"type\": \"float\", \"min\": 0.0, \"max\": 1.0, \"default\": 0.2},\n",
    "    \"ensemble\" : {\"type\": \"integer\", \"default\": 0},\n",
    "    \"ensemble_id\" : {\"type\": \"integer\", \"default\": 0}\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "eUuupPuDbVNi"
   },
   "source": [
    "We will also want to set-up some elements that will stay the same for all the examples. These components are the encoder - which is responsible for creating input files for our simulation and the decoder - which is responsible for parsing the output of the simulation."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "UX-nxATzcquI"
   },
   "source": [
    "For the Encoder we use the GenericEncoder class. It is a very simple template based encoder. It finds a specified delimiter, and replaces the variable name that follows that delimiter with the corresponding value. In our case the template file looks like follows:\n",
    "\n",
    "```\n",
    "{\n",
    "    \"grid_size\" : $grid_size,\n",
    "    \"n\" : $n,\n",
    "    \"duration\" : $duration,\n",
    "    \"mortality\" : $mortality\n",
    "}\n",
    "```\n",
    "\n",
    "From this template, a JSON file will be created and then passed to the simulation script as an argument. EasyVVUQ has other encoders as well. For example the [Jinja encoder](https://easyvvuq.readthedocs.io/en/dev/source/easyvvuq.encoders.html#module-easyvvuq.encoders.jinja_encoder).\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:34.852846Z",
     "start_time": "2021-06-09T11:12:34.851526Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "3MUHaMR5chT-"
   },
   "outputs": [],
   "source": [
    "encoder = uq.encoders.GenericEncoder(\n",
    "    template_fname='epidemic/epidemic.template',\n",
    "    delimiter='$',\n",
    "    target_filename='epidemic_in.json')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ht9C4JYmoK_k"
   },
   "source": [
    "Since the quantity of interest (number of turns until the disease is erradicated) is a function of the simulation output (it is the iteration number of the last row) we need to extend the Decoder class to take this in to account. To this end we inherit from SimpleCSV decoder and redefine the `parse_sim_output` method to take the last value of the `iteration` column in the file produced by the simulation. This gives us the length in turns for which the simulation ran or in other words before the disease disappeared in our simulation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:34.855087Z",
     "start_time": "2021-06-09T11:12:34.853524Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "kloY-UH2my9q"
   },
   "outputs": [],
   "source": [
    "class EpidemicDecoder(uq.decoders.SimpleCSV):   # , decoder_name='epidemic_decoder'\n",
    "    def parse_sim_output(self, run_info={}):\n",
    "        result = super().parse_sim_output(run_info)\n",
    "        return {'iteration': result['iteration'][-1]}\n",
    "\n",
    "decoder = EpidemicDecoder(\n",
    "    target_filename=\"output.csv\", output_columns=[\"iteration\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "p9QHSWzva6td"
   },
   "source": [
    "We will also define a helper function that will execute the simulation with the provided input files. This function takes a campaign object, creates the directories with input files and then runs our script in them with those files as inputs. The exact details of this process can be found [here](https://easyvvuq.readthedocs.io/en/dev/source/easyvvuq.html#easyvvuq.campaign.Campaign.populate_runs_dir) and [here](https://easyvvuq.readthedocs.io/en/dev/source/easyvvuq.html#easyvvuq.campaign.Campaign.apply_for_each_run_dir)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "uhwgG1Wc4tIj"
   },
   "source": [
    "## Basic Example\n",
    "\n",
    "We start by creating an\n",
    "EasyVVUQ Campaign. Here we call it 'epidemic_basic'. :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:34.880753Z",
     "start_time": "2021-06-09T11:12:34.855742Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "Cf5cFQXr4tIj"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "db_location = sqlite:///./epidemic_basicw4x3b3o2/campaign.db\n",
      "active_sampler_id = None\n",
      "campaign_name = epidemic_basic\n",
      "campaign_dir = /Volumes/Samsung8TB/dpc/GIT/EasyVVUQ/tutorials/./epidemic_basicw4x3b3o2\n",
      "campaign_id = 1\n",
      "\n"
     ]
    }
   ],
   "source": [
    "execute = ExecuteLocal('{}/epidemic/epidemic.py epidemic_in.json output.csv'.format(os.getcwd()))\n",
    "\n",
    "actions = Actions(CreateRunDirectory('/tmp'), Encode(encoder), execute, Decode(decoder))\n",
    "\n",
    "campaign = uq.Campaign(name='epidemic_basic', params=params, actions=actions)\n",
    "print(campaign)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "IxyJdZD6oOLd"
   },
   "source": [
    "We then want to describe our application. This means passing parameter dictionary, enoder, decoder and collater to the campaign object."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "EBUKUTyL4tIz"
   },
   "source": [
    "For this particular task we are not interested in the relationship between input parameters and the behavior of the simulation. All we want is to see how much the result varies between runs that are identical but for the random seed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:34.886374Z",
     "start_time": "2021-06-09T11:12:34.881426Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "1eDGvW3J4tI2"
   },
   "outputs": [],
   "source": [
    "from easyvvuq.sampling import EmptySampler\n",
    "campaign.set_sampler(EmptySampler())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "syDEBtKzo7TS"
   },
   "source": [
    "EmptySampler is a convenience class for such cases. However, another option is, if your simulation provides the option to specify different seeds, to draw the seeds from a probability distribution. In cases like these you could specify the sampler like this:\n",
    "\n",
    "```\n",
    "from easyvvuq.sampling import RandomSampler\n",
    "\n",
    "vary = {'seed' : cp.DiscreteUniform(0, MAX_SEED)}\n",
    "sampler = RandomSampler(vary)\n",
    "```\n",
    "\n",
    "In the above example `MAX_SEED` is the maximum value the seed can take plus one.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "E5RS1S0Hp97R"
   },
   "source": [
    "We now want to execute the simulations and collate the results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.281533Z",
     "start_time": "2021-06-09T11:12:34.902648Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "VvHnsBKXqAqX"
   },
   "outputs": [],
   "source": [
    "campaign.execute(nsamples=20).collate()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "2QARvB3rH7_U"
   },
   "source": [
    "We can now see what the result is. It will be a DataFrame containing the number of iterations before the disease is erradicated. Calling collate on the campaign object will find and parse the output files of our simulation and then produce a single DataFrame with the results.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.292752Z",
     "start_time": "2021-06-09T11:12:42.282448Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "o4Dy4R1bH08q"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   run_id iteration grid_size   n duration mortality ensemble\n",
      "        0         0         0   0        0         0        0\n",
      "0       1      73.0        10  20       28       0.2        0\n",
      "1       2      80.0        10  20       28       0.2        0\n",
      "2       3      89.0        10  20       28       0.2        0\n",
      "3       4      73.0        10  20       28       0.2        0\n",
      "4       5      57.0        10  20       28       0.2        0\n",
      "5       6      88.0        10  20       28       0.2        0\n",
      "6       7      72.0        10  20       28       0.2        0\n",
      "7       8      81.0        10  20       28       0.2        0\n",
      "8       9      81.0        10  20       28       0.2        0\n",
      "9      10      67.0        10  20       28       0.2        0\n",
      "10     11      53.0        10  20       28       0.2        0\n",
      "11     12      59.0        10  20       28       0.2        0\n",
      "12     13      73.0        10  20       28       0.2        0\n",
      "13     14      55.0        10  20       28       0.2        0\n",
      "14     15      56.0        10  20       28       0.2        0\n",
      "15     16      61.0        10  20       28       0.2        0\n",
      "16     17      65.0        10  20       28       0.2        0\n",
      "17     18      70.0        10  20       28       0.2        0\n",
      "18     19      61.0        10  20       28       0.2        0\n",
      "19     20      75.0        10  20       28       0.2        0\n",
      "20     21      77.0        10  20       28       0.2        0\n",
      "21     22      75.0        10  20       28       0.2        0\n",
      "22     23      71.0        10  20       28       0.2        0\n",
      "23     24      76.0        10  20       28       0.2        0\n",
      "24     25      58.0        10  20       28       0.2        0\n",
      "25     26      71.0        10  20       28       0.2        0\n",
      "26     27      89.0        10  20       28       0.2        0\n",
      "27     28      89.0        10  20       28       0.2        0\n",
      "28     29      77.0        10  20       28       0.2        0\n",
      "29     30      65.0        10  20       28       0.2        0\n",
      "30     31      76.0        10  20       28       0.2        0\n",
      "31     32      71.0        10  20       28       0.2        0\n",
      "32     33      85.0        10  20       28       0.2        0\n",
      "33     34     101.0        10  20       28       0.2        0\n",
      "34     35      65.0        10  20       28       0.2        0\n",
      "35     36      88.0        10  20       28       0.2        0\n",
      "36     37      64.0        10  20       28       0.2        0\n",
      "37     38      75.0        10  20       28       0.2        0\n",
      "38     39      27.0        10  20       28       0.2        0\n",
      "39     40      56.0        10  20       28       0.2        0\n"
     ]
    }
   ],
   "source": [
    "df = campaign.get_collation_result()\n",
    "print(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "yJLFypP34tJD"
   },
   "source": [
    "This collated data is stored in the campaign database. An analysis\n",
    "element, here EnsembleBoot (which performs [bootstraping](https://en.wikipedia.org/wiki/Bootstrapping_(statistics))), can then be applied to the campaign's\n",
    "collation result. :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.363816Z",
     "start_time": "2021-06-09T11:12:42.293555Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "CacIi3Ky4tJE"
   },
   "outputs": [],
   "source": [
    "analysis = uq.analysis.EnsembleBoot(qoi_cols=[(\"iteration\", 0)], stat_func=np.mean, stat_name='mean', alpha=0.05)\n",
    "campaign.apply_analysis(analysis)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "rGgiN7cA4tJF"
   },
   "source": [
    "The output of this is dependent on the type of analysis element. :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.367749Z",
     "start_time": "2021-06-09T11:12:42.364534Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "-TvW8CIj4tJG"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     iteration                      \n",
      "             0                      \n",
      "          mean        low       high\n",
      "True     71.35  66.974375  75.075625\n"
     ]
    }
   ],
   "source": [
    "# Get Descriptive Statistics\n",
    "results = campaign.get_last_analysis()\n",
    "print(results)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "fDPk1Eahzf8v"
   },
   "source": [
    "The above gives the mean value of the number of iterations before the disease eradication and the 95% confidence region for that estimator."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "PAWMNIaZWciH"
   },
   "source": [
    "# Parameter Sweep\n",
    "\n",
    "Suppose we want to examine the behaviour of the model by doing a sweep across a range of parameter values. Let's say that given a fixed population size and fixed mortality rate we want to see how does the model behaves if we vary the disease duration parameter from 4 days to, for example, 28 days. We will do this in 2 day increments. Also, in order to perform some kind of statistical analysis of the results we will want to do sample the same parameter set several times. We will do this using the replicas mechanism in the EasyVVUQ."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Yyd_LYTZoWro"
   },
   "source": [
    "First we define a new campaign."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:29:46.118661Z",
     "start_time": "2021-06-09T11:29:46.071961Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "-PTFAsd1YExO"
   },
   "outputs": [],
   "source": [
    "campaign = uq.Campaign(name='epidemic_sweep', params=params, actions=actions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "gtqE6rVVovw5"
   },
   "source": [
    "We need to define a different kind of sampler. In our case we want BasicSweep which allows one to sample an n-dimensional grid of values. In this case we specify a single parameter and give it a list of values. If you add more parameters the sampler will then sample a Cartesian product of those lists. For examples `{'a' : [1, 2], 'b' : [3, 4]}` will result in `{'a': 1, 'b': 3}`, `{'a': 1, 'b': 4}`, `{'a': 2, 'b': 3}` and `{'a': 2, 'b': 4}` being sampled."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:29:49.221449Z",
     "start_time": "2021-06-09T11:29:49.217536Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "QYga_8MMXkb3"
   },
   "outputs": [],
   "source": [
    "sweep = {\n",
    "    \"duration\" : list(range(2, 30, 2))\n",
    "}\n",
    "sweep_sampler = uq.sampling.BasicSweep(sweep=sweep)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will also want to sample each point multiple times in order to construct confidence intervals. To this end we use a ReplicaSampler element which wraps around our `sweep_sampler`. In essence it will simply run the simulation with the same inputs multiple times with different random number generator seeds."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:29:52.540073Z",
     "start_time": "2021-06-09T11:29:52.524509Z"
    }
   },
   "outputs": [],
   "source": [
    "from easyvvuq.sampling import ReplicaSampler\n",
    "sampler = ReplicaSampler(sweep_sampler)\n",
    "campaign.set_sampler(sampler)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "nVF9fqai8iU6"
   },
   "source": [
    "The following steps are the same as in our more basic example. Note that execution might take some time. That is because we are running 280 simulations on the machine that is hosting this notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:34:03.542242Z",
     "start_time": "2021-06-09T11:30:40.889363Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "zomhNA23Y6iD"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 744/744 [03:20<00:00,  3.70it/s]\n"
     ]
    }
   ],
   "source": [
    "campaign.execute(nsamples=20 * 14).collate(progress_bar=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:40:37.497524Z",
     "start_time": "2021-06-09T11:40:37.464994Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "hF6BuutSmF4a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    run_id iteration duration ensemble_id grid_size   n mortality ensemble\n",
      "         0         0        0           0         0   0         0        0\n",
      "0        1       2.0        2           0        10  20       0.2        0\n",
      "1        2       3.0        4           1        10  20       0.2        0\n",
      "2        3       8.0        6           2        10  20       0.2        0\n",
      "3        4       7.0        8           3        10  20       0.2        0\n",
      "4        5       9.0       10           4        10  20       0.2        0\n",
      "..     ...       ...      ...         ...       ...  ..       ...      ...\n",
      "835    836      85.0       20           9        10  20       0.2        0\n",
      "836    837      70.0       22          10        10  20       0.2        0\n",
      "837    838      74.0       24          11        10  20       0.2        0\n",
      "838    839      72.0       26          12        10  20       0.2        0\n",
      "839    840      71.0       28          13        10  20       0.2        0\n",
      "\n",
      "[840 rows x 8 columns]\n"
     ]
    }
   ],
   "source": [
    "df = campaign.get_collation_result()\n",
    "print(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "_fesrbXQ9sX2"
   },
   "source": [
    "One important difference in the analysis stage below is the addition of the `groupby` keyword argument. This means, essentially, that bootstrapping will be done on rows with the same `ensemble_id` value separately. So in the end we will get `(30 - 2) / 2 = 14` triples of the mean value and lower/upper confidence bounds."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:41:44.665046Z",
     "start_time": "2021-06-09T11:41:43.908106Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "mWQ2WEb6sTUm"
   },
   "outputs": [],
   "source": [
    "analysis = uq.analysis.EnsembleBoot(qoi_cols=[(\"iteration\", 0)], groupby=('ensemble_id', 0), stat_func=np.mean, stat_name='mean')\n",
    "campaign.apply_analysis(analysis)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "KjOTJzW9-cUc"
   },
   "source": [
    "We can now view the resulting `DataFrame`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:42:39.642684Z",
     "start_time": "2021-06-09T11:42:39.633347Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "jRVhGK84sXcM"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                  iteration                      \n",
      "                          0                      \n",
      "                       mean        low       high\n",
      "(ensemble_id, 0)                                 \n",
      "0                  1.333333   1.183333   1.500417\n",
      "1                  3.950000   3.600000   4.350000\n",
      "2                  9.283333   8.082917  10.767500\n",
      "3                 14.666667  12.850000  16.866667\n",
      "4                 28.800000  25.733333  32.384167\n",
      "5                 33.850000  30.048750  37.917083\n",
      "6                 43.283333  38.098750  47.784167\n",
      "7                 51.150000  46.316250  55.767500\n",
      "8                 52.583333  48.648750  56.183333\n",
      "9                 64.525000  60.316250  68.450833\n",
      "10                64.975000  61.815417  68.134583\n",
      "11                65.808333  62.082917  69.600417\n",
      "12                71.500000  68.265417  74.800417\n",
      "13                75.658333  72.150000  79.435000\n"
     ]
    }
   ],
   "source": [
    "results = campaign.get_last_analysis()\n",
    "print(results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:45:45.308509Z",
     "start_time": "2021-06-09T11:45:45.297426Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "LwWwIsNbszAJ"
   },
   "outputs": [],
   "source": [
    "#plt.errorbar(list(range(2, 30, 2)), \n",
    "#             results[('iteration', 0)]['mean'].values,\n",
    "#             np.array([results['iteration']['mean'].values - results['iteration']['low'].values,\n",
    "#                       results['iteration']['high'].values - results['iteration']['mean'].values]))\n",
    "#plt.xlabel('duration (days)')\n",
    "#plt.ylabel('days until eradication')\n",
    "#plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ElXBj81rUDpu"
   },
   "source": [
    "# Remote Execution"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This part of the tutorial assumes that you have a Kubernetes cluster access configured. In other words the `~/.kube/config` file needs to be populated with the data that Kubernetes API can use to connect to a cluster. The exact details for how to do this will depend on your cloud service provider. For example, in order to start a cluster on GKE I would need to run the following command (or similar):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.595747Z",
     "start_time": "2021-06-09T11:12:29.810Z"
    }
   },
   "outputs": [],
   "source": [
    "!gcloud container clusters create easyvvuq"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To use this functionality you would first need to create a Docker image for your simulation. There are many resource on how to do this. See for example [here](https://docs.docker.com/get-started/part2/). For your convenience below is the Dockerfile we have used when creating an EasyVVUQ image that will be used in his example. You also need to publish your Docker image to where it can be downloaded by the Kubernetes cluster. An obvious place is [Docker Hub](https://hub.docker.com/).\n",
    "\n",
    "```\n",
    "FROM ubuntu:latest\n",
    "\n",
    "RUN apt-get update && \\\n",
    "    apt-get install -y python3-pip && \\\n",
    "    apt-get install -y git && \\\n",
    "    apt-get install -y tini && \\\n",
    "    pip3 install easyvvuq && \\\n",
    "    git clone https://github.com/UCL-CCS/EasyVVUQ.git\n",
    "\n",
    "ENTRYPOINT [\"tini\", \"--\"]\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After this, the only other bit of boring admin you need to do is to create a Kubernetes pod configuration. Again here is the one that was used in this tutorial:\n",
    "\n",
    "```\n",
    "apiVersion: v1\n",
    "kind: Pod\n",
    "metadata:\n",
    "  name: epidemic\n",
    "spec:\n",
    "  restartPolicy: Never\n",
    "  containers:\n",
    "  - name: epidemic\n",
    "    image: orbitfold/easyvvuq:latest\n",
    "    command: [\"/bin/sh\", \"-c\"]\n",
    "    args: [\"python3 /EasyVVUQ/docs/epidemic/epidemic.py /config/example.json out.csv && cat out.csv\"]\n",
    "```\n",
    "\n",
    "It will likely be the same for your simulation. Please note the image name, and the command that will be used to execute the simulation. In the current implementation your simulation needs to put all output to the standard output. So after creating the output CSV file we use `cat` to print it to the screen. Your way of doing this is likely to be different."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have, so far, drawn 20 samples in our parameter sweep which has resulted in fairly wide error bars. We can shrink them by adding more samples. To this end will draw 80 more samples and create the corresponding input files."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.596133Z",
     "start_time": "2021-06-09T11:12:29.819Z"
    }
   },
   "outputs": [],
   "source": [
    "campaign.draw_samples(60 * 14)\n",
    "campaign.populate_runs_dir()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we call `apply_for_each_run_dir` with the `ExecuteKubernetes` action. This will submit the jobs to the Kubernetes cluster and the execution will automatically start."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.596589Z",
     "start_time": "2021-06-09T11:12:29.820Z"
    }
   },
   "outputs": [],
   "source": [
    "statuses = campaign.apply_for_each_run_dir(\n",
    "      uq.actions.ExecuteKubernetes('../tutorials/kubernetes/epidemic.yaml', \n",
    "                                   ['epidemic_in.json'], 'output.csv'), batch_size=8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.597012Z",
     "start_time": "2021-06-09T11:12:29.820Z"
    }
   },
   "outputs": [],
   "source": [
    "statuses = campaign.sample_and_apply(\n",
    "    60 * 14,\n",
    "    uq.actions.ExecuteKubernetes(\n",
    "        '../tutorials/kubernetes/epidemic.yaml', \n",
    "        ['epidemic_in.json'], \n",
    "        'output.csv'), 8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.597567Z",
     "start_time": "2021-06-09T11:12:29.820Z"
    }
   },
   "outputs": [],
   "source": [
    "statuses.start()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.598018Z",
     "start_time": "2021-06-09T11:12:29.821Z"
    }
   },
   "outputs": [],
   "source": [
    "print(statuses.progress())\n",
    "statuses.actions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.598422Z",
     "start_time": "2021-06-09T11:12:29.821Z"
    }
   },
   "outputs": [],
   "source": [
    "campaign.collate()\n",
    "df = campaign.get_collation_result()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.598868Z",
     "start_time": "2021-06-09T11:12:29.821Z"
    }
   },
   "outputs": [],
   "source": [
    "analysis = uq.analysis.EnsembleBoot(qoi_cols=[\"iteration\"], groupby='ensemble', stat_func=np.mean, stat_name='mean')\n",
    "campaign.apply_analysis(analysis)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.599293Z",
     "start_time": "2021-06-09T11:12:29.821Z"
    }
   },
   "outputs": [],
   "source": [
    "results = campaign.get_last_analysis()\n",
    "print(results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:12:42.599735Z",
     "start_time": "2021-06-09T11:12:29.822Z"
    }
   },
   "outputs": [],
   "source": [
    "plt.errorbar(list(range(2, 30, 2)), \n",
    "             results['iteration']['mean'].values,\n",
    "             np.array([results['iteration']['mean'].values - results['iteration']['low'].values,\n",
    "                       results['iteration']['high'].values - results['iteration']['mean'].values]))\n",
    "plt.xlabel('duration (days)')\n",
    "plt.ylabel('days until eradication')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sensitivy Analysis with QMC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:46:35.815632Z",
     "start_time": "2021-06-09T11:46:35.779662Z"
    }
   },
   "outputs": [],
   "source": [
    "campaign = uq.Campaign(name='sobol_method', params=params, actions=actions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:47:01.688736Z",
     "start_time": "2021-06-09T11:47:01.660143Z"
    }
   },
   "outputs": [],
   "source": [
    "vary = {\n",
    "    \"duration\": cp.DiscreteUniform(7, 14),\n",
    "    \"mortality\": cp.Uniform(0.1, 0.3),\n",
    "}\n",
    "qmc_sampler = uq.sampling.QMCSampler(vary, 128)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:47:05.484714Z",
     "start_time": "2021-06-09T11:47:05.465185Z"
    }
   },
   "outputs": [],
   "source": [
    "campaign.set_sampler(qmc_sampler)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:49:47.023377Z",
     "start_time": "2021-06-09T11:48:06.147543Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 512/512 [01:40<00:00,  5.09it/s]\n"
     ]
    }
   ],
   "source": [
    "campaign.execute().collate(progress_bar=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:49:54.180092Z",
     "start_time": "2021-06-09T11:49:54.139812Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    run_id iteration   duration mortality grid_size   n ensemble ensemble_id\n",
      "         0         0          0         0         0   0        0           0\n",
      "0        1      13.0   6.511719  0.175293        10  20        0           0\n",
      "1        2      14.0  10.082031  0.175293        10  20        0           0\n",
      "2        3      12.0   6.511719  0.197363        10  20        0           0\n",
      "3        4       9.0  10.082031  0.197363        10  20        0           0\n",
      "4        5      24.0  10.511719  0.275293        10  20        0           0\n",
      "..     ...       ...        ...       ...       ...  ..      ...         ...\n",
      "507    508      13.0  10.394531  0.173926        10  20        0           0\n",
      "508    509      12.0   6.574219  0.258105        10  20        0           0\n",
      "509    510      55.0  14.394531  0.258105        10  20        0           0\n",
      "510    511       5.0   6.574219  0.273926        10  20        0           0\n",
      "511    512      24.0  14.394531  0.273926        10  20        0           0\n",
      "\n",
      "[512 rows x 8 columns]\n"
     ]
    }
   ],
   "source": [
    "df = campaign.get_collation_result()\n",
    "print(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:51:00.317005Z",
     "start_time": "2021-06-09T11:50:58.505493Z"
    }
   },
   "outputs": [],
   "source": [
    "from easyvvuq.analysis import QMCAnalysis\n",
    "analysis = QMCAnalysis(qmc_sampler, ['n', 'duration', 'mortality'])\n",
    "campaign.apply_analysis(analysis)\n",
    "results = campaign.get_last_analysis()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:57:02.446459Z",
     "start_time": "2021-06-09T11:57:02.441366Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'n': {'duration': array([0.]), 'mortality': array([0.])},\n",
       " 'duration': {'duration': array([1.01554484]), 'mortality': array([0.])},\n",
       " 'mortality': {'duration': array([0.]), 'mortality': array([1.01224875])}}"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results.sobols_first()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:57:20.647435Z",
     "start_time": "2021-06-09T11:57:20.643067Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'duration': array([0.]), 'mortality': array([1.01224875])}"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results.sobols_first('mortality')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T11:57:35.652593Z",
     "start_time": "2021-06-09T11:57:35.645974Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'duration': array([1.01554484]), 'mortality': array([0.])}"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results.sobols_first('duration')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T12:17:21.596591Z",
     "start_time": "2021-06-09T12:17:21.581676Z"
    }
   },
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>n</th>\n",
       "      <th>duration</th>\n",
       "      <th>mortality</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>0</th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>20.0</td>\n",
       "      <td>10.484375</td>\n",
       "      <td>0.199609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>var</th>\n",
       "      <td>0.0</td>\n",
       "      <td>5.333023</td>\n",
       "      <td>0.003333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2.309334</td>\n",
       "      <td>0.057733</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         n   duration mortality\n",
       "         0          0         0\n",
       "mean  20.0  10.484375  0.199609\n",
       "var    0.0   5.333023  0.003333\n",
       "std    0.0   2.309334  0.057733"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "xdGqgdlg4tJI"
   },
   "source": [
    "# Using Stochastic Collocation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T12:01:04.182632Z",
     "start_time": "2021-06-09T12:01:04.114377Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "eZJDR7_ZvUTx"
   },
   "outputs": [],
   "source": [
    "campaign = uq.Campaign(name='epidemic_sc', params=params, actions=actions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T12:01:07.432934Z",
     "start_time": "2021-06-09T12:01:06.849040Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "HocadTwF4tJJ"
   },
   "outputs": [],
   "source": [
    "vary = {\n",
    "    \"n\": cp.DiscreteUniform(10, 100),\n",
    "    \"duration\": cp.DiscreteUniform(7, 56),\n",
    "    \"mortality\": cp.Uniform(0.0, 1.0),\n",
    "}\n",
    "sampler = uq.sampling.SCSampler(vary=vary, polynomial_order=9)\n",
    "campaign.set_sampler(sampler)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T12:05:38.606297Z",
     "start_time": "2021-06-09T12:01:45.686430Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "E6LLEaUaN12F"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [03:52<00:00,  4.30it/s]\n"
     ]
    }
   ],
   "source": [
    "campaign.execute().collate(progress_bar=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T12:08:48.259099Z",
     "start_time": "2021-06-09T12:08:48.175980Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "hdlG-EpXOARP"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    run_id iteration     n duration mortality grid_size ensemble ensemble_id\n",
      "         0         0     0        0         0         0        0           0\n",
      "0        1      12.0  14.0      9.0  0.013047        10        0           0\n",
      "1        2      10.0  14.0      9.0  0.067468        10        0           0\n",
      "2        3      22.0  14.0      9.0  0.160295        10        0           0\n",
      "3        4      13.0  14.0      9.0  0.283302        10        0           0\n",
      "4        5       8.0  14.0      9.0  0.425563        10        0           0\n",
      "..     ...       ...   ...      ...       ...       ...      ...         ...\n",
      "995    996      69.0  96.0     54.0  0.574437        10        0           0\n",
      "996    997      71.0  96.0     54.0  0.716698        10        0           0\n",
      "997    998      74.0  96.0     54.0  0.839705        10        0           0\n",
      "998    999      65.0  96.0     54.0  0.932532        10        0           0\n",
      "999   1000      69.0  96.0     54.0  0.986953        10        0           0\n",
      "\n",
      "[1000 rows x 8 columns]\n"
     ]
    }
   ],
   "source": [
    "df = campaign.get_collation_result()\n",
    "print(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "qrmRhOxB4tJN"
   },
   "source": [
    "And the analysis can be done with: :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T12:10:23.720134Z",
     "start_time": "2021-06-09T12:10:23.007088Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "c-dD72LA4tJO"
   },
   "outputs": [],
   "source": [
    "analysis = uq.analysis.SCAnalysis(sampler=sampler, qoi_cols=[\"iteration\"])\n",
    "campaign.apply_analysis(analysis)\n",
    "results_sc = campaign.get_last_analysis()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T12:17:59.935340Z",
     "start_time": "2021-06-09T12:17:59.923391Z"
    }
   },
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>iteration</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>62.304993</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>24.803309</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>var</th>\n",
       "      <td>615.204157</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       iteration\n",
       "               0\n",
       "mean   62.304993\n",
       "std    24.803309\n",
       "var   615.204157"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_sc.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T12:13:04.848224Z",
     "start_time": "2021-06-09T12:13:04.771831Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([133196.93383935])"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "analysis.surrogate('iteration', np.array([5, 7, 0.7]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T12:15:08.230696Z",
     "start_time": "2021-06-09T12:14:53.828510Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "P0BcqazCnb9x"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n = np.linspace(5, 100, 20)\n",
    "duration = np.linspace(7, 50, 20)\n",
    "mortality = 0.7\n",
    "grid = np.meshgrid(n, duration, mortality)\n",
    "z = np.array([analysis.surrogate('iteration', np.array([grid[0].flatten()[i], grid[1].flatten()[i], mortality])) for i in range(n.shape[0] * duration.shape[0])]).flatten()\n",
    "fig = plt.figure(figsize=(10, 10), dpi= 80, facecolor='w', edgecolor='k')\n",
    "ax = fig.gca(projection='3d')\n",
    "ax.set_zlim(0, 200)\n",
    "ax.plot_trisurf(grid[0].flatten(), grid[1].flatten(), z)\n",
    "ax.set_xlabel('n')\n",
    "ax.set_ylabel('duration')\n",
    "ax.set_zlabel('time')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "-ph5PGWyweNr"
   },
   "source": [
    "Let us quickly check the correspondence of the surrogate model to the data we have collected during our parameter sweep experiment. What we're looking for is the curve should be within the confidence interval we have calculated."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T12:16:05.723232Z",
     "start_time": "2021-06-09T12:16:05.156571Z"
    },
    "colab": {},
    "colab_type": "code",
    "id": "KkR5FHvx1BAw"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "n = [20] * 14\n",
    "duration = list(range(2, 30, 2))\n",
    "mortality = [0.2] * 14\n",
    "\n",
    "plt.plot(([analysis.surrogate('iteration', np.array([n_, duration_, mortality_]))[0] for n_, duration_, mortality_ in zip(n, duration, mortality)]))\n",
    "plt.ylim(0, 100)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "EasyVVUQ Tutorial with a Toy Epidemiological Model",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.9.4"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autoclose": false,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}