{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "R_get_started.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyPkRcZlL+l67RshCOJXMEYs",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "ir",
      "display_name": "R"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/lisatwyw/data-gym/blob/master/R_get_started.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MJCtIAZLTnfA",
        "colab_type": "text"
      },
      "source": [
        "## A) Mount to google drive## "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rCEtur7fajGn",
        "colab_type": "code",
        "outputId": "f5f34c55-d21c-4d5e-857f-1b3bcfbfa15a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 109
        }
      },
      "source": [
        "cat(system('python3 -c \"from google.colab import drive\\ndrive.mount()\"', intern=TRUE), sep='\\n', wait=TRUE)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Warning message in system(\"python3 -c \\\"from google.colab import drive\\ndrive.mount()\\\"\", :\n",
            "“running command 'python3 -c \"from google.colab import drive\n",
            "drive.mount()\"' had status 1”\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "TRUE\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T6aH0TahbghZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "?system"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ezwFft-tSSUW",
        "colab_type": "text"
      },
      "source": [
        "## B) Try out comorbidity calculation ##\n",
        "\n",
        "- import devtools\n",
        "- downloand from github\n",
        "- import comorbidity library"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ipk2vXhuSL0Q",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "install.packages(\"devtools\")\n",
        "require(devtools)\n",
        "\n",
        "install_github( 'ellessenne/comorbidity' )\n",
        "\n",
        "# load the comorbidity package\n",
        "library(comorbidity)\n",
        "# set a seed for reproducibility\n",
        "set.seed(1)\n",
        "\n",
        "# simulate 50 ICD-10 codes for 5 individuals\n",
        "x <- data.frame(\n",
        "  id = sample(1:5, size = 50, replace = TRUE),\n",
        "  code = sample_diag(n = 50),\n",
        "  stringsAsFactors = FALSE\n",
        ")\n",
        "x <- x[order(x$id, x$code), ]\n",
        "print(head(x, n = 15), row.names = FALSE)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e2jONWnvS3Lq",
        "colab_type": "text"
      },
      "source": [
        "\n",
        "*** All of below is under construction!***\n",
        "\n",
        "## c) EHR data ##"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bcsRVS2UF7Wx",
        "colab_type": "code",
        "outputId": "ed52e96b-5b5a-46a0-edbc-52161143dcb1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 109
        }
      },
      "source": [
        "url='https://github.com/rOpenHealth/rEHR/blob/master/data-raw/example_tables.R'\n",
        "\n",
        "install.packages('bit')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Installing package into ‘/usr/local/lib/R/site-library’\n",
            "(as ‘lib’ is unspecified)\n",
            "\n",
            "Warning message in install.packages(\"bit\"):\n",
            "“installation of package ‘bit’ had non-zero exit status”\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Z0dvjeOS4GJN",
        "colab_type": "code",
        "outputId": "dfbcd32d-cb39-4d6b-9f97-d86c8883ab75",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 668
        }
      },
      "source": [
        "remotes::install_github(\"rOpenHealth/rEHR\")\n",
        "require(rEHR)\n",
        "\n",
        "## Use the simulated ehr files supplied with the package to build our database\n",
        "ehr_path <- dirname(system.file(\"ehr_data\", \"ehr_Clinical.txt\", package = \"rEHR\"))\n",
        "## create a new database connection to a temporary file\n",
        "db <- database(tempfile(fileext = \".sqlite\"))\n",
        "## Import multiple data files into the database\n",
        "import_CPRD_data(db, data_dir = ehr_path,\n",
        "filetypes = c(\"Clinical\", \"Consultation\",\n",
        "\"Patient\", \"Practice\",\n",
        "\"Referral\"),\n",
        "dateformat = \"%Y-%m-%d\",\n",
        "yob_origin = 1800,\n",
        "regex = \"ehr\",\n",
        "recursive = TRUE)\n",
        "## Individual files can also be added:\n",
        "add_to_database(db, files = system.file(\"ehr_data\", \"ehr_Therapy.txt\", package = \"rEHR\"),\n",
        "table_name = \"Therapy\", dateformat = \"%Y-%m-%d\")\n",
        "## Use the overloaded `head` function to view a list of\n",
        "## tables or the head of individual tables:\n",
        "head(db)\n",
        "\n",
        "url='https://github.com/rOpenHealth/rEHR/blob/master/data-raw/example_tables.R'\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading GitHub repo rOpenHealth/rEHR@master\n",
            "\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "sqldf    (NA    -> 0.4-11  ) [CRAN]\n",
            "RSQLite  (NA    -> 2.2.0   ) [CRAN]\n",
            "xlsx     (NA    -> 0.6.3   ) [CRAN]\n",
            "bit64    (NA    -> 0.9-7   ) [CRAN]\n",
            "bit      (NA    -> 1.1-15.2) [CRAN]\n",
            "rJava    (NA    -> 0.9-12  ) [CRAN]\n",
            "gsubfn   (NA    -> 0.7     ) [CRAN]\n",
            "chron    (NA    -> 2.3-55  ) [CRAN]\n",
            "tibble   (2.1.3 -> 3.0.0   ) [CRAN]\n",
            "xlsxjars (NA    -> 0.6.1   ) [CRAN]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Installing 10 packages: sqldf, RSQLite, xlsx, bit64, bit, rJava, gsubfn, chron, tibble, xlsxjars\n",
            "\n",
            "Installing packages into ‘/usr/local/lib/R/site-library’\n",
            "(as ‘lib’ is unspecified)\n",
            "\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "error",
          "ename": "ERROR",
          "evalue": "ignored",
          "traceback": [
            "Error: Failed to install 'rEHR' from GitHub:\n  (converted from warning) installation of package ‘bit’ had non-zero exit status\nTraceback:\n",
            "1. remotes::install_github(\"rOpenHealth/rEHR\")",
            "2. install_remotes(remotes, auth_token = auth_token, host = host, \n .     dependencies = dependencies, upgrade = upgrade, force = force, \n .     quiet = quiet, build = build, build_opts = build_opts, build_manual = build_manual, \n .     build_vignettes = build_vignettes, repos = repos, type = type, \n .     ...)",
            "3. tryCatch(res[[i]] <- install_remote(remotes[[i]], ...), error = function(e) {\n .     stop(remote_install_error(remotes[[i]], e))\n . })",
            "4. tryCatchList(expr, classes, parentenv, handlers)",
            "5. tryCatchOne(expr, names, parentenv, handlers[[1L]])",
            "6. value[[3L]](cond)"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YrDGP56mSx1v",
        "colab_type": "text"
      },
      "source": [
        "## D) Query EHR records ##"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xt3AO5sEAlAl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "diabetes_codes <- clinical_codes[clinical_codes$list == \"Diabetes\",]\n",
        "\n",
        "select_events(db, tab = \"Clinical\", columns = c(\"patid\", \"eventdate\", \"medcode\"), where = \"medcode %in% .(diabetes_codes$medcode) & eventdate < '2006-01-01' & eventdate >= '2005-01-01'\")\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lePkHRF0Sqr9",
        "colab_type": "text"
      },
      "source": [
        "## E) Eurostat ##"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NQ8LYoyVq7dF",
        "colab_type": "code",
        "outputId": "d642339d-1a4f-4aed-c71c-412500d90171",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 503
        }
      },
      "source": [
        "install.packages(\"eurostat\")\n",
        "require(eurostat)\n",
        "\n",
        "library(eurostat)\n",
        "\n",
        "toc <- eurostat::get_eurostat_toc()\n",
        "\n",
        "df <- fetch_healthdata(resource=\"hosp\", filter=list(addr_city=\"SAN FRANCISCO\"))\n",
        "\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Installing package into ‘/usr/local/lib/R/site-library’\n",
            "(as ‘lib’ is unspecified)\n",
            "\n",
            "also installing the dependencies ‘e1071’, ‘units’, ‘classInt’, ‘sf’\n",
            "\n",
            "\n",
            "Warning message in install.packages(\"eurostat\"):\n",
            "“installation of package ‘e1071’ had non-zero exit status”\n",
            "Warning message in install.packages(\"eurostat\"):\n",
            "“installation of package ‘units’ had non-zero exit status”\n",
            "Warning message in install.packages(\"eurostat\"):\n",
            "“installation of package ‘classInt’ had non-zero exit status”\n",
            "Warning message in install.packages(\"eurostat\"):\n",
            "“installation of package ‘sf’ had non-zero exit status”\n",
            "Warning message in install.packages(\"eurostat\"):\n",
            "“installation of package ‘eurostat’ had non-zero exit status”\n",
            "Loading required package: eurostat\n",
            "\n",
            "Warning message in library(package, lib.loc = lib.loc, character.only = TRUE, logical.return = TRUE, :\n",
            "“there is no package called ‘eurostat’”\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "error",
          "ename": "ERROR",
          "evalue": "ignored",
          "traceback": [
            "Error in library(eurostat): there is no package called ‘eurostat’\nTraceback:\n",
            "1. library(eurostat)"
          ]
        }
      ]
    }
  ]
}