sklearn_classifier/sklearn_classifier.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Generic scikit-learn classifier\n",
    "\n",
    "run any scikit-learn compatible classifier or list of classifiers\n",
    "### **Function steps** :\n",
    "> **Generate a scikit-learn model configuration** using the `model_pkg_class` parameter\n",
    ">   * input a package and class name, for example, `sklearn.linear_model.LogisticRegression`  \n",
    ">   * mlrun will find the class and instantiate a copy using default parameters  \n",
    ">   * You can modify both the model class instantiator and the fit methods  (other functions could be similarly modified)\n",
    "\n",
    "> **Get a sample of data** from a data source\n",
    ">  * select all rows using -1\n",
    ">  * select a random sample of rows using a negative integer\n",
    ">  * select consecutive rows using a positive integer\n",
    "\n",
    "> **Split the data** into train, validation, and test sets \n",
    ">  * the test set is saved as an artifact and never seen again until testing\n",
    ">  * WIP: this will be parametrized to produce cross-validator splits (one way of performing CV)\n",
    "\n",
    "> **Train the model** using xgboost in one of its flavours (dask, gpu, mpi...)\n",
    "\n",
    "> **Pickle / serialize the model**\n",
    ">  * models can be pickled or saved as json (a newer xgboost format)\n",
    "\n",
    "> **Evaluate the model**\n",
    ">  * a custom evaluator can be provided, see function doc for details"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Example steps**\n",
    "\n",
    "1. [Data exploration](#Data-exploration)\n",
    "2. [Importing the function](#Importing-the-function)\n",
    "3. [Setup sklearn trainer parameters](#Setup-sklearn-trainer-parameters)\n",
    "4. [Set model parameters and run locally](#Set-model-parameters-and-run-locally)\n",
    "5. [Set model parameters and run remotely](#Set-model-parameters-and-run-remotely)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Data exploration**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ".. _iris_dataset:\n",
      "\n",
      "Iris plants dataset\n",
      "--------------------\n",
      "\n",
      "**Data Set Characteristics:**\n",
      "\n",
      "    :Number of Instances: 150 (50 in each of three classes)\n",
      "    :Number of Attributes: 4 numeric, predictive attributes and the class\n",
      "    :Attribute Information:\n",
      "        - sepal length in cm\n",
      "        - sepal width in cm\n",
      "        - petal length in cm\n",
      "        - petal width in cm\n",
      "        - class:\n",
      "                - Iris-Setosa\n",
      "                - Iris-Versicolour\n",
      "                - Iris-Virginica\n",
      "                \n",
      "    :Summary Statistics:\n",
      "\n",
      "    ============== ==== ==== ======= ===== ====================\n",
      "                    Min  Max   Mean    SD   Class Correlation\n",
      "    ============== ==== ==== ======= ===== ====================\n",
      "    sepal length:   4.3  7.9   5.84   0.83    0.7826\n",
      "    sepal width:    2.0  4.4   3.05   0.43   -0.4194\n",
      "    petal length:   1.0  6.9   3.76   1.76    0.9490  (high!)\n",
      "    petal width:    0.1  2.5   1.20   0.76    0.9565  (high!)\n",
      "    ============== ==== ==== ======= ===== ====================\n",
      "\n",
      "    :Missing Attribute Values: None\n",
      "    :Class Distribution: 33.3% for each of 3 classes.\n",
      "    :Creator: R.A. Fisher\n",
      "    :Donor: Michael Marshall (MARSHALL%PLU@io.arc.nasa.gov)\n",
      "    :Date: July, 1988\n",
      "\n",
      "The famous Iris database, first used by Sir R.A. Fisher. The dataset is taken\n",
      "from Fisher's paper. Note that it's the same as in R, but not as in the UCI\n",
      "Machine Learning Repository, which has two wrong data points.\n",
      "\n",
      "This is perhaps the best known database to be found in the\n",
      "pattern recognition literature.  Fisher's paper is a classic in the field and\n",
      "is referenced frequently to this day.  (See Duda & Hart, for example.)  The\n",
      "data set contains 3 classes of 50 instances each, where each class refers to a\n",
      "type of iris plant.  One class is linearly separable from the other 2; the\n",
      "latter are NOT linearly separable from each other.\n",
      "\n",
      ".. topic:: References\n",
      "\n",
      "   - Fisher, R.A. \"The use of multiple measurements in taxonomic problems\"\n",
      "     Annual Eugenics, 7, Part II, 179-188 (1936); also in \"Contributions to\n",
      "     Mathematical Statistics\" (John Wiley, NY, 1950).\n",
      "   - Duda, R.O., & Hart, P.E. (1973) Pattern Classification and Scene Analysis.\n",
      "     (Q327.D83) John Wiley & Sons.  ISBN 0-471-22361-1.  See page 218.\n",
      "   - Dasarathy, B.V. (1980) \"Nosing Around the Neighborhood: A New System\n",
      "     Structure and Classification Rule for Recognition in Partially Exposed\n",
      "     Environments\".  IEEE Transactions on Pattern Analysis and Machine\n",
      "     Intelligence, Vol. PAMI-2, No. 1, 67-71.\n",
      "   - Gates, G.W. (1972) \"The Reduced Nearest Neighbor Rule\".  IEEE Transactions\n",
      "     on Information Theory, May 1972, 431-433.\n",
      "   - See also: 1988 MLC Proceedings, 54-64.  Cheeseman et al\"s AUTOCLASS II\n",
      "     conceptual clustering system finds 3 classes in the data.\n",
      "   - Many, many more ...\n"
     ]
    }
   ],
   "source": [
    "# Dataset is generated from sklearn's load_iris\n",
    "from sklearn.datasets import load_iris\n",
    "\n",
    "data = load_iris()\n",
    "print(data['DESCR'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "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>sepal length (cm)</th>\n",
       "      <th>sepal width (cm)</th>\n",
       "      <th>petal length (cm)</th>\n",
       "      <th>petal width (cm)</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.1</td>\n",
       "      <td>3.5</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4.7</td>\n",
       "      <td>3.2</td>\n",
       "      <td>1.3</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.6</td>\n",
       "      <td>3.1</td>\n",
       "      <td>1.5</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>3.6</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sepal length (cm)  sepal width (cm)  petal length (cm)  petal width (cm)  \\\n",
       "0                5.1               3.5                1.4               0.2   \n",
       "1                4.9               3.0                1.4               0.2   \n",
       "2                4.7               3.2                1.3               0.2   \n",
       "3                4.6               3.1                1.5               0.2   \n",
       "4                5.0               3.6                1.4               0.2   \n",
       "\n",
       "   label  \n",
       "0      0  \n",
       "1      0  \n",
       "2      0  \n",
       "3      0  \n",
       "4      0  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "data_url = 'https://s3.wasabisys.com/iguazio/data/function-marketplace-data/sklearn_classifier/iris_dataset.csv'\n",
    "df = pd.read_csv(data_url)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Importing the function**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "> 2021-10-17 11:44:40,760 [info] loaded project function-marketplace from MLRun DB\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<mlrun.runtimes.kubejob.KubejobRuntime at 0x7f9bd28f8490>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import mlrun\n",
    "mlrun.set_environment(project='function-marketplace')\n",
    "\n",
    "# Importing the function from the hub\n",
    "fn = mlrun.import_function(\"hub://sklearn_classifier\")\n",
    "fn.apply(mlrun.auto_mount())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "### **Setup sklearn trainer parameters**\n",
    "\n",
    "the following task paramaters are common to all runs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "task_params = {\n",
    "    #\"name\" : \"sklearn classifier\",\n",
    "    \"params\" : {\n",
    "        \"sample\"             : -1,\n",
    "        \"test_size\"          : 0.30,\n",
    "        \"train_val_split\"    : 0.70,\n",
    "        \"random_state\"       : 1,\n",
    "        \"n_jobs\"             : -1,\n",
    "        \"plots_dest\"         : \"plots-p\",\n",
    "        \"models_dest\"        : 'sklearn-clfmodel'}}\n",
    "\n",
    "models = [\n",
    "    \"sklearn.ensemble.RandomForestClassifier\", \n",
    "    \"sklearn.linear_model.LogisticRegression\",\n",
    "    \"sklearn.ensemble.AdaBoostClassifier\"\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "### **Set model parameters and run locally**\n",
    "\n",
    "* loop over a list of candidate models, update the task and run a local trainer for that model  \n",
    "* optionally customize some parameters for each model\n",
    "* accumulate the resulting artifacts in a list of dicts for debugging and testing purposes  \n",
    "\n",
    "please note that the following parameters have been chosen without regard to optimization and model tuning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "> 2021-10-17 11:44:41,149 [info] starting run sklearn_ensemble_RandomForestClassifier uid=5410f6bc8d9a4273820a440134f0dc07 DB=http://mlrun-api:8080\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>\n",
       ".dictlist {\n",
       "  background-color: #4EC64B;\n",
       "  text-align: center;\n",
       "  margin: 4px;\n",
       "  border-radius: 3px; padding: 0px 3px 1px 3px; display: inline-block;}\n",
       ".artifact {\n",
       "  cursor: pointer;\n",
       "  background-color: #4EC64B;\n",
       "  text-align: left;\n",
       "  margin: 4px; border-radius: 3px; padding: 0px 3px 1px 3px; display: inline-block;\n",
       "}\n",
       "div.block.hidden {\n",
       "  display: none;\n",
       "}\n",
       ".clickable {\n",
       "  cursor: pointer;\n",
       "}\n",
       ".ellipsis {\n",
       "  display: inline-block;\n",
       "  max-width: 60px;\n",
       "  white-space: nowrap;\n",
       "  overflow: hidden;\n",
       "  text-overflow: ellipsis;\n",
       "}\n",
       ".master-wrapper {\n",
       "  display: flex;\n",
       "  flex-flow: row nowrap;\n",
       "  justify-content: flex-start;\n",
       "  align-items: stretch;\n",
       "}\n",
       ".master-tbl {\n",
       "  flex: 3\n",
       "}\n",
       ".master-wrapper > div {\n",
       "  margin: 4px;\n",
       "  padding: 10px;\n",
       "}\n",
       "iframe.fileview {\n",
       "  border: 0 none;\n",
       "  height: 100%;\n",
       "  width: 100%;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       ".pane-header-title {\n",
       "  width: 80%;\n",
       "  font-weight: 500;\n",
       "}\n",
       ".pane-header {\n",
       "  line-height: 1;\n",
       "  background-color: #4EC64B;\n",
       "  padding: 3px;\n",
       "}\n",
       ".pane-header .close {\n",
       "  font-size: 20px;\n",
       "  font-weight: 700;\n",
       "  float: right;\n",
       "  margin-top: -5px;\n",
       "}\n",
       ".master-wrapper .right-pane {\n",
       "  border: 1px inset silver;\n",
       "  width: 40%;\n",
       "  min-height: 300px;\n",
       "  flex: 3\n",
       "  min-width: 500px;\n",
       "}\n",
       ".master-wrapper * {\n",
       "  box-sizing: border-box;\n",
       "}\n",
       "</style><script>\n",
       "function copyToClipboard(fld) {\n",
       "    if (document.queryCommandSupported && document.queryCommandSupported('copy')) {\n",
       "        var textarea = document.createElement('textarea');\n",
       "        textarea.textContent = fld.innerHTML;\n",
       "        textarea.style.position = 'fixed';\n",
       "        document.body.appendChild(textarea);\n",
       "        textarea.select();\n",
       "\n",
       "        try {\n",
       "            return document.execCommand('copy'); // Security exception may be thrown by some browsers.\n",
       "        } catch (ex) {\n",
       "\n",
       "        } finally {\n",
       "            document.body.removeChild(textarea);\n",
       "        }\n",
       "    }\n",
       "}\n",
       "function expandPanel(el) {\n",
       "  const panelName = \"#\" + el.getAttribute('paneName');\n",
       "  console.log(el.title);\n",
       "\n",
       "  document.querySelector(panelName + \"-title\").innerHTML = el.title\n",
       "  iframe = document.querySelector(panelName + \"-body\");\n",
       "\n",
       "  const tblcss = `<style> body { font-family: Arial, Helvetica, sans-serif;}\n",
       "    #csv { margin-bottom: 15px; }\n",
       "    #csv table { border-collapse: collapse;}\n",
       "    #csv table td { padding: 4px 8px; border: 1px solid silver;} </style>`;\n",
       "\n",
       "  function csvToHtmlTable(str) {\n",
       "    return '<div id=\"csv\"><table><tr><td>' +  str.replace(/[\\n\\r]+$/g, '').replace(/[\\n\\r]+/g, '</td></tr><tr><td>')\n",
       "      .replace(/,/g, '</td><td>') + '</td></tr></table></div>';\n",
       "  }\n",
       "\n",
       "  function reqListener () {\n",
       "    if (el.title.endsWith(\".csv\")) {\n",
       "      iframe.setAttribute(\"srcdoc\", tblcss + csvToHtmlTable(this.responseText));\n",
       "    } else {\n",
       "      iframe.setAttribute(\"srcdoc\", this.responseText);\n",
       "    }\n",
       "    console.log(this.responseText);\n",
       "  }\n",
       "\n",
       "  const oReq = new XMLHttpRequest();\n",
       "  oReq.addEventListener(\"load\", reqListener);\n",
       "  oReq.open(\"GET\", el.title);\n",
       "  oReq.send();\n",
       "\n",
       "\n",
       "  //iframe.src = el.title;\n",
       "  const resultPane = document.querySelector(panelName + \"-pane\");\n",
       "  if (resultPane.classList.contains(\"hidden\")) {\n",
       "    resultPane.classList.remove(\"hidden\");\n",
       "  }\n",
       "}\n",
       "function closePanel(el) {\n",
       "  const panelName = \"#\" + el.getAttribute('paneName')\n",
       "  const resultPane = document.querySelector(panelName + \"-pane\");\n",
       "  if (!resultPane.classList.contains(\"hidden\")) {\n",
       "    resultPane.classList.add(\"hidden\");\n",
       "  }\n",
       "}\n",
       "\n",
       "</script>\n",
       "<div class=\"master-wrapper\">\n",
       "  <div class=\"block master-tbl\"><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>project</th>\n",
       "      <th>uid</th>\n",
       "      <th>iter</th>\n",
       "      <th>start</th>\n",
       "      <th>state</th>\n",
       "      <th>name</th>\n",
       "      <th>labels</th>\n",
       "      <th>inputs</th>\n",
       "      <th>parameters</th>\n",
       "      <th>results</th>\n",
       "      <th>artifacts</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>function-marketplace</td>\n",
       "      <td><div title=\"5410f6bc8d9a4273820a440134f0dc07\"><a href=\"https://dashboard.default-tenant.app.dev39.lab.iguazeng.com/mlprojects/function-marketplace/jobs/monitor/5410f6bc8d9a4273820a440134f0dc07/overview\" target=\"_blank\" >...34f0dc07</a></div></td>\n",
       "      <td>0</td>\n",
       "      <td>Oct 17 11:44:41</td>\n",
       "      <td>completed</td>\n",
       "      <td>sklearn_ensemble_RandomForestClassifier</td>\n",
       "      <td><div class=\"dictlist\">v3io_user=dani</div><div class=\"dictlist\">kind=</div><div class=\"dictlist\">owner=dani</div><div class=\"dictlist\">host=jupyter-dani-6bfbd76d96-zxx6f</div><div class=\"dictlist\">class=sklearn.ensemble.RandomForestClassifier</div></td>\n",
       "      <td><div title=\"https://s3.wasabisys.com/iguazio/data/iris/iris_dataset.csv\">dataset</div></td>\n",
       "      <td><div class=\"dictlist\">model_pkg_class=sklearn.ensemble.RandomForestClassifier</div><div class=\"dictlist\">label_column=label</div><div class=\"dictlist\">CLASS_max_depth=5</div></td>\n",
       "      <td><div class=\"dictlist\">accuracy=0.90625</div><div class=\"dictlist\">test-error=0.09375</div><div class=\"dictlist\">auc-micro=0.99462890625</div><div class=\"dictlist\">auc-weighted=1.0</div><div class=\"dictlist\">f1-score=0.886039886039886</div><div class=\"dictlist\">precision_score=0.875</div><div class=\"dictlist\">recall_score=0.9333333333333332</div></td>\n",
       "      <td><div title=\"v3io:///projects/function-marketplace/artifacts/data/test_set.parquet\">test_set</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result3bafa864\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/confusion-matrix.html\">confusion-matrix</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result3bafa864\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/feature-importances.html\">feature-importances</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result3bafa864\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/precision-recall-multiclass.html\">precision-recall-multiclass</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result3bafa864\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/roc-multiclass.html\">roc-multiclass</div><div title=\"v3io:///projects/function-marketplace/artifacts/model/\">model</div></td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div></div>\n",
       "  <div id=\"result3bafa864-pane\" class=\"right-pane block hidden\">\n",
       "    <div class=\"pane-header\">\n",
       "      <span id=\"result3bafa864-title\" class=\"pane-header-title\">Title</span>\n",
       "      <span onclick=\"closePanel(this)\" paneName=\"result3bafa864\" class=\"close clickable\">&times;</span>\n",
       "    </div>\n",
       "    <iframe class=\"fileview\" id=\"result3bafa864-body\"></iframe>\n",
       "  </div>\n",
       "</div>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<b> > to track results use the .show() or .logs() methods  or <a href=\"https://dashboard.default-tenant.app.dev39.lab.iguazeng.com/mlprojects/function-marketplace/jobs/monitor/5410f6bc8d9a4273820a440134f0dc07/overview\" target=\"_blank\">click here</a> to open in UI</b>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "> 2021-10-17 11:44:44,151 [info] run executed, status=completed\n",
      "> 2021-10-17 11:44:44,154 [info] starting run sklearn_linear_model_LogisticRegression uid=0326c1bd8f3d4f7a95719e6e51ff1483 DB=http://mlrun-api:8080\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>\n",
       ".dictlist {\n",
       "  background-color: #4EC64B;\n",
       "  text-align: center;\n",
       "  margin: 4px;\n",
       "  border-radius: 3px; padding: 0px 3px 1px 3px; display: inline-block;}\n",
       ".artifact {\n",
       "  cursor: pointer;\n",
       "  background-color: #4EC64B;\n",
       "  text-align: left;\n",
       "  margin: 4px; border-radius: 3px; padding: 0px 3px 1px 3px; display: inline-block;\n",
       "}\n",
       "div.block.hidden {\n",
       "  display: none;\n",
       "}\n",
       ".clickable {\n",
       "  cursor: pointer;\n",
       "}\n",
       ".ellipsis {\n",
       "  display: inline-block;\n",
       "  max-width: 60px;\n",
       "  white-space: nowrap;\n",
       "  overflow: hidden;\n",
       "  text-overflow: ellipsis;\n",
       "}\n",
       ".master-wrapper {\n",
       "  display: flex;\n",
       "  flex-flow: row nowrap;\n",
       "  justify-content: flex-start;\n",
       "  align-items: stretch;\n",
       "}\n",
       ".master-tbl {\n",
       "  flex: 3\n",
       "}\n",
       ".master-wrapper > div {\n",
       "  margin: 4px;\n",
       "  padding: 10px;\n",
       "}\n",
       "iframe.fileview {\n",
       "  border: 0 none;\n",
       "  height: 100%;\n",
       "  width: 100%;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       ".pane-header-title {\n",
       "  width: 80%;\n",
       "  font-weight: 500;\n",
       "}\n",
       ".pane-header {\n",
       "  line-height: 1;\n",
       "  background-color: #4EC64B;\n",
       "  padding: 3px;\n",
       "}\n",
       ".pane-header .close {\n",
       "  font-size: 20px;\n",
       "  font-weight: 700;\n",
       "  float: right;\n",
       "  margin-top: -5px;\n",
       "}\n",
       ".master-wrapper .right-pane {\n",
       "  border: 1px inset silver;\n",
       "  width: 40%;\n",
       "  min-height: 300px;\n",
       "  flex: 3\n",
       "  min-width: 500px;\n",
       "}\n",
       ".master-wrapper * {\n",
       "  box-sizing: border-box;\n",
       "}\n",
       "</style><script>\n",
       "function copyToClipboard(fld) {\n",
       "    if (document.queryCommandSupported && document.queryCommandSupported('copy')) {\n",
       "        var textarea = document.createElement('textarea');\n",
       "        textarea.textContent = fld.innerHTML;\n",
       "        textarea.style.position = 'fixed';\n",
       "        document.body.appendChild(textarea);\n",
       "        textarea.select();\n",
       "\n",
       "        try {\n",
       "            return document.execCommand('copy'); // Security exception may be thrown by some browsers.\n",
       "        } catch (ex) {\n",
       "\n",
       "        } finally {\n",
       "            document.body.removeChild(textarea);\n",
       "        }\n",
       "    }\n",
       "}\n",
       "function expandPanel(el) {\n",
       "  const panelName = \"#\" + el.getAttribute('paneName');\n",
       "  console.log(el.title);\n",
       "\n",
       "  document.querySelector(panelName + \"-title\").innerHTML = el.title\n",
       "  iframe = document.querySelector(panelName + \"-body\");\n",
       "\n",
       "  const tblcss = `<style> body { font-family: Arial, Helvetica, sans-serif;}\n",
       "    #csv { margin-bottom: 15px; }\n",
       "    #csv table { border-collapse: collapse;}\n",
       "    #csv table td { padding: 4px 8px; border: 1px solid silver;} </style>`;\n",
       "\n",
       "  function csvToHtmlTable(str) {\n",
       "    return '<div id=\"csv\"><table><tr><td>' +  str.replace(/[\\n\\r]+$/g, '').replace(/[\\n\\r]+/g, '</td></tr><tr><td>')\n",
       "      .replace(/,/g, '</td><td>') + '</td></tr></table></div>';\n",
       "  }\n",
       "\n",
       "  function reqListener () {\n",
       "    if (el.title.endsWith(\".csv\")) {\n",
       "      iframe.setAttribute(\"srcdoc\", tblcss + csvToHtmlTable(this.responseText));\n",
       "    } else {\n",
       "      iframe.setAttribute(\"srcdoc\", this.responseText);\n",
       "    }\n",
       "    console.log(this.responseText);\n",
       "  }\n",
       "\n",
       "  const oReq = new XMLHttpRequest();\n",
       "  oReq.addEventListener(\"load\", reqListener);\n",
       "  oReq.open(\"GET\", el.title);\n",
       "  oReq.send();\n",
       "\n",
       "\n",
       "  //iframe.src = el.title;\n",
       "  const resultPane = document.querySelector(panelName + \"-pane\");\n",
       "  if (resultPane.classList.contains(\"hidden\")) {\n",
       "    resultPane.classList.remove(\"hidden\");\n",
       "  }\n",
       "}\n",
       "function closePanel(el) {\n",
       "  const panelName = \"#\" + el.getAttribute('paneName')\n",
       "  const resultPane = document.querySelector(panelName + \"-pane\");\n",
       "  if (!resultPane.classList.contains(\"hidden\")) {\n",
       "    resultPane.classList.add(\"hidden\");\n",
       "  }\n",
       "}\n",
       "\n",
       "</script>\n",
       "<div class=\"master-wrapper\">\n",
       "  <div class=\"block master-tbl\"><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>project</th>\n",
       "      <th>uid</th>\n",
       "      <th>iter</th>\n",
       "      <th>start</th>\n",
       "      <th>state</th>\n",
       "      <th>name</th>\n",
       "      <th>labels</th>\n",
       "      <th>inputs</th>\n",
       "      <th>parameters</th>\n",
       "      <th>results</th>\n",
       "      <th>artifacts</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>function-marketplace</td>\n",
       "      <td><div title=\"0326c1bd8f3d4f7a95719e6e51ff1483\"><a href=\"https://dashboard.default-tenant.app.dev39.lab.iguazeng.com/mlprojects/function-marketplace/jobs/monitor/0326c1bd8f3d4f7a95719e6e51ff1483/overview\" target=\"_blank\" >...51ff1483</a></div></td>\n",
       "      <td>0</td>\n",
       "      <td>Oct 17 11:44:44</td>\n",
       "      <td>completed</td>\n",
       "      <td>sklearn_linear_model_LogisticRegression</td>\n",
       "      <td><div class=\"dictlist\">v3io_user=dani</div><div class=\"dictlist\">kind=</div><div class=\"dictlist\">owner=dani</div><div class=\"dictlist\">host=jupyter-dani-6bfbd76d96-zxx6f</div><div class=\"dictlist\">class=sklearn.linear_model.LogisticRegression</div></td>\n",
       "      <td><div title=\"https://s3.wasabisys.com/iguazio/data/iris/iris_dataset.csv\">dataset</div></td>\n",
       "      <td><div class=\"dictlist\">model_pkg_class=sklearn.linear_model.LogisticRegression</div><div class=\"dictlist\">label_column=label</div><div class=\"dictlist\">CLASS_solver=liblinear</div></td>\n",
       "      <td><div class=\"dictlist\">accuracy=0.875</div><div class=\"dictlist\">test-error=0.125</div><div class=\"dictlist\">auc-micro=0.97265625</div><div class=\"dictlist\">auc-weighted=0.997004357298475</div><div class=\"dictlist\">f1-score=0.8534798534798534</div><div class=\"dictlist\">precision_score=0.8518518518518517</div><div class=\"dictlist\">recall_score=0.9111111111111111</div></td>\n",
       "      <td><div title=\"v3io:///projects/function-marketplace/artifacts/data/test_set.parquet\">test_set</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result29c7da0f\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/confusion-matrix.html\">confusion-matrix</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result29c7da0f\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/precision-recall-multiclass.html\">precision-recall-multiclass</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result29c7da0f\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/roc-multiclass.html\">roc-multiclass</div><div title=\"v3io:///projects/function-marketplace/artifacts/model/\">model</div></td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div></div>\n",
       "  <div id=\"result29c7da0f-pane\" class=\"right-pane block hidden\">\n",
       "    <div class=\"pane-header\">\n",
       "      <span id=\"result29c7da0f-title\" class=\"pane-header-title\">Title</span>\n",
       "      <span onclick=\"closePanel(this)\" paneName=\"result29c7da0f\" class=\"close clickable\">&times;</span>\n",
       "    </div>\n",
       "    <iframe class=\"fileview\" id=\"result29c7da0f-body\"></iframe>\n",
       "  </div>\n",
       "</div>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<b> > to track results use the .show() or .logs() methods  or <a href=\"https://dashboard.default-tenant.app.dev39.lab.iguazeng.com/mlprojects/function-marketplace/jobs/monitor/0326c1bd8f3d4f7a95719e6e51ff1483/overview\" target=\"_blank\">click here</a> to open in UI</b>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "> 2021-10-17 11:44:45,759 [info] run executed, status=completed\n",
      "> 2021-10-17 11:44:45,762 [info] starting run sklearn_ensemble_AdaBoostClassifier uid=d5386ba6222e41f8a9f46d96ba866bc0 DB=http://mlrun-api:8080\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>\n",
       ".dictlist {\n",
       "  background-color: #4EC64B;\n",
       "  text-align: center;\n",
       "  margin: 4px;\n",
       "  border-radius: 3px; padding: 0px 3px 1px 3px; display: inline-block;}\n",
       ".artifact {\n",
       "  cursor: pointer;\n",
       "  background-color: #4EC64B;\n",
       "  text-align: left;\n",
       "  margin: 4px; border-radius: 3px; padding: 0px 3px 1px 3px; display: inline-block;\n",
       "}\n",
       "div.block.hidden {\n",
       "  display: none;\n",
       "}\n",
       ".clickable {\n",
       "  cursor: pointer;\n",
       "}\n",
       ".ellipsis {\n",
       "  display: inline-block;\n",
       "  max-width: 60px;\n",
       "  white-space: nowrap;\n",
       "  overflow: hidden;\n",
       "  text-overflow: ellipsis;\n",
       "}\n",
       ".master-wrapper {\n",
       "  display: flex;\n",
       "  flex-flow: row nowrap;\n",
       "  justify-content: flex-start;\n",
       "  align-items: stretch;\n",
       "}\n",
       ".master-tbl {\n",
       "  flex: 3\n",
       "}\n",
       ".master-wrapper > div {\n",
       "  margin: 4px;\n",
       "  padding: 10px;\n",
       "}\n",
       "iframe.fileview {\n",
       "  border: 0 none;\n",
       "  height: 100%;\n",
       "  width: 100%;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       ".pane-header-title {\n",
       "  width: 80%;\n",
       "  font-weight: 500;\n",
       "}\n",
       ".pane-header {\n",
       "  line-height: 1;\n",
       "  background-color: #4EC64B;\n",
       "  padding: 3px;\n",
       "}\n",
       ".pane-header .close {\n",
       "  font-size: 20px;\n",
       "  font-weight: 700;\n",
       "  float: right;\n",
       "  margin-top: -5px;\n",
       "}\n",
       ".master-wrapper .right-pane {\n",
       "  border: 1px inset silver;\n",
       "  width: 40%;\n",
       "  min-height: 300px;\n",
       "  flex: 3\n",
       "  min-width: 500px;\n",
       "}\n",
       ".master-wrapper * {\n",
       "  box-sizing: border-box;\n",
       "}\n",
       "</style><script>\n",
       "function copyToClipboard(fld) {\n",
       "    if (document.queryCommandSupported && document.queryCommandSupported('copy')) {\n",
       "        var textarea = document.createElement('textarea');\n",
       "        textarea.textContent = fld.innerHTML;\n",
       "        textarea.style.position = 'fixed';\n",
       "        document.body.appendChild(textarea);\n",
       "        textarea.select();\n",
       "\n",
       "        try {\n",
       "            return document.execCommand('copy'); // Security exception may be thrown by some browsers.\n",
       "        } catch (ex) {\n",
       "\n",
       "        } finally {\n",
       "            document.body.removeChild(textarea);\n",
       "        }\n",
       "    }\n",
       "}\n",
       "function expandPanel(el) {\n",
       "  const panelName = \"#\" + el.getAttribute('paneName');\n",
       "  console.log(el.title);\n",
       "\n",
       "  document.querySelector(panelName + \"-title\").innerHTML = el.title\n",
       "  iframe = document.querySelector(panelName + \"-body\");\n",
       "\n",
       "  const tblcss = `<style> body { font-family: Arial, Helvetica, sans-serif;}\n",
       "    #csv { margin-bottom: 15px; }\n",
       "    #csv table { border-collapse: collapse;}\n",
       "    #csv table td { padding: 4px 8px; border: 1px solid silver;} </style>`;\n",
       "\n",
       "  function csvToHtmlTable(str) {\n",
       "    return '<div id=\"csv\"><table><tr><td>' +  str.replace(/[\\n\\r]+$/g, '').replace(/[\\n\\r]+/g, '</td></tr><tr><td>')\n",
       "      .replace(/,/g, '</td><td>') + '</td></tr></table></div>';\n",
       "  }\n",
       "\n",
       "  function reqListener () {\n",
       "    if (el.title.endsWith(\".csv\")) {\n",
       "      iframe.setAttribute(\"srcdoc\", tblcss + csvToHtmlTable(this.responseText));\n",
       "    } else {\n",
       "      iframe.setAttribute(\"srcdoc\", this.responseText);\n",
       "    }\n",
       "    console.log(this.responseText);\n",
       "  }\n",
       "\n",
       "  const oReq = new XMLHttpRequest();\n",
       "  oReq.addEventListener(\"load\", reqListener);\n",
       "  oReq.open(\"GET\", el.title);\n",
       "  oReq.send();\n",
       "\n",
       "\n",
       "  //iframe.src = el.title;\n",
       "  const resultPane = document.querySelector(panelName + \"-pane\");\n",
       "  if (resultPane.classList.contains(\"hidden\")) {\n",
       "    resultPane.classList.remove(\"hidden\");\n",
       "  }\n",
       "}\n",
       "function closePanel(el) {\n",
       "  const panelName = \"#\" + el.getAttribute('paneName')\n",
       "  const resultPane = document.querySelector(panelName + \"-pane\");\n",
       "  if (!resultPane.classList.contains(\"hidden\")) {\n",
       "    resultPane.classList.add(\"hidden\");\n",
       "  }\n",
       "}\n",
       "\n",
       "</script>\n",
       "<div class=\"master-wrapper\">\n",
       "  <div class=\"block master-tbl\"><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>project</th>\n",
       "      <th>uid</th>\n",
       "      <th>iter</th>\n",
       "      <th>start</th>\n",
       "      <th>state</th>\n",
       "      <th>name</th>\n",
       "      <th>labels</th>\n",
       "      <th>inputs</th>\n",
       "      <th>parameters</th>\n",
       "      <th>results</th>\n",
       "      <th>artifacts</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>function-marketplace</td>\n",
       "      <td><div title=\"d5386ba6222e41f8a9f46d96ba866bc0\"><a href=\"https://dashboard.default-tenant.app.dev39.lab.iguazeng.com/mlprojects/function-marketplace/jobs/monitor/d5386ba6222e41f8a9f46d96ba866bc0/overview\" target=\"_blank\" >...ba866bc0</a></div></td>\n",
       "      <td>0</td>\n",
       "      <td>Oct 17 11:44:45</td>\n",
       "      <td>completed</td>\n",
       "      <td>sklearn_ensemble_AdaBoostClassifier</td>\n",
       "      <td><div class=\"dictlist\">v3io_user=dani</div><div class=\"dictlist\">kind=</div><div class=\"dictlist\">owner=dani</div><div class=\"dictlist\">host=jupyter-dani-6bfbd76d96-zxx6f</div><div class=\"dictlist\">class=sklearn.ensemble.AdaBoostClassifier</div></td>\n",
       "      <td><div title=\"https://s3.wasabisys.com/iguazio/data/iris/iris_dataset.csv\">dataset</div></td>\n",
       "      <td><div class=\"dictlist\">model_pkg_class=sklearn.ensemble.AdaBoostClassifier</div><div class=\"dictlist\">label_column=label</div><div class=\"dictlist\">CLASS_n_estimators=200</div><div class=\"dictlist\">CLASS_learning_rate=0.01</div></td>\n",
       "      <td><div class=\"dictlist\">accuracy=0.90625</div><div class=\"dictlist\">test-error=0.09375</div><div class=\"dictlist\">auc-micro=0.96044921875</div><div class=\"dictlist\">auc-weighted=1.0</div><div class=\"dictlist\">f1-score=0.886039886039886</div><div class=\"dictlist\">precision_score=0.875</div><div class=\"dictlist\">recall_score=0.9333333333333332</div></td>\n",
       "      <td><div title=\"v3io:///projects/function-marketplace/artifacts/data/test_set.parquet\">test_set</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result2a4365e3\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/confusion-matrix.html\">confusion-matrix</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result2a4365e3\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/feature-importances.html\">feature-importances</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result2a4365e3\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/precision-recall-multiclass.html\">precision-recall-multiclass</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result2a4365e3\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/roc-multiclass.html\">roc-multiclass</div><div title=\"v3io:///projects/function-marketplace/artifacts/model/\">model</div></td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div></div>\n",
       "  <div id=\"result2a4365e3-pane\" class=\"right-pane block hidden\">\n",
       "    <div class=\"pane-header\">\n",
       "      <span id=\"result2a4365e3-title\" class=\"pane-header-title\">Title</span>\n",
       "      <span onclick=\"closePanel(this)\" paneName=\"result2a4365e3\" class=\"close clickable\">&times;</span>\n",
       "    </div>\n",
       "    <iframe class=\"fileview\" id=\"result2a4365e3-body\"></iframe>\n",
       "  </div>\n",
       "</div>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<b> > to track results use the .show() or .logs() methods  or <a href=\"https://dashboard.default-tenant.app.dev39.lab.iguazeng.com/mlprojects/function-marketplace/jobs/monitor/d5386ba6222e41f8a9f46d96ba866bc0/overview\" target=\"_blank\">click here</a> to open in UI</b>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "> 2021-10-17 11:44:48,144 [info] run executed, status=completed\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "outputs = []\n",
    "for model in models:\n",
    "    task_copy = task_params.copy()\n",
    "    task_copy.update(\n",
    "        {\n",
    "            \"params\":{ \"model_pkg_class\" : model,\n",
    "                       \"label_column\"    : \"label\"}\n",
    "        }\n",
    "    )\n",
    "    \n",
    "    # customize specific model parameters\n",
    "    if \"RandomForestClassifier\" in model:\n",
    "        task_copy[\"params\"].update({\"CLASS_max_depth\" : 5})\n",
    "\n",
    "    if \"LogisticRegression\" in model:\n",
    "        task_copy[\"params\"].update({\"CLASS_solver\" : \"liblinear\"})\n",
    "    \n",
    "    if \"AdaBoostClassifier\" in model:\n",
    "        task_copy[\"params\"].update({\"CLASS_n_estimators\"  : 200,\n",
    "                                    \"CLASS_learning_rate\" : 0.01\n",
    "                                   })\n",
    "    \n",
    "    name = model.replace('.', '_')\n",
    "    output = fn.run(mlrun.new_task(**task_copy),\n",
    "                    name=name,\n",
    "                    inputs={\"dataset\"  : data_url},\n",
    "                    local=True)\n",
    "    \n",
    "    outputs.append({name: output.outputs})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'sklearn_ensemble_RandomForestClassifier': {'accuracy': 0.90625,\n",
       "   'test-error': 0.09375,\n",
       "   'auc-micro': 0.99462890625,\n",
       "   'auc-weighted': 1.0,\n",
       "   'f1-score': 0.886039886039886,\n",
       "   'precision_score': 0.875,\n",
       "   'recall_score': 0.9333333333333332,\n",
       "   'test_set': 'store://artifacts/function-marketplace/sklearn_ensemble_RandomForestClassifier_test_set:5410f6bc8d9a4273820a440134f0dc07',\n",
       "   'confusion-matrix': 'v3io:///projects/function-marketplace/artifacts/model/plots/confusion-matrix.html',\n",
       "   'feature-importances': 'v3io:///projects/function-marketplace/artifacts/model/plots/feature-importances.html',\n",
       "   'precision-recall-multiclass': 'v3io:///projects/function-marketplace/artifacts/model/plots/precision-recall-multiclass.html',\n",
       "   'roc-multiclass': 'v3io:///projects/function-marketplace/artifacts/model/plots/roc-multiclass.html',\n",
       "   'model': 'store://artifacts/function-marketplace/sklearn_ensemble_RandomForestClassifier_model:5410f6bc8d9a4273820a440134f0dc07'}},\n",
       " {'sklearn_linear_model_LogisticRegression': {'accuracy': 0.875,\n",
       "   'test-error': 0.125,\n",
       "   'auc-micro': 0.97265625,\n",
       "   'auc-weighted': 0.997004357298475,\n",
       "   'f1-score': 0.8534798534798534,\n",
       "   'precision_score': 0.8518518518518517,\n",
       "   'recall_score': 0.9111111111111111,\n",
       "   'test_set': 'store://artifacts/function-marketplace/sklearn_linear_model_LogisticRegression_test_set:0326c1bd8f3d4f7a95719e6e51ff1483',\n",
       "   'confusion-matrix': 'v3io:///projects/function-marketplace/artifacts/model/plots/confusion-matrix.html',\n",
       "   'precision-recall-multiclass': 'v3io:///projects/function-marketplace/artifacts/model/plots/precision-recall-multiclass.html',\n",
       "   'roc-multiclass': 'v3io:///projects/function-marketplace/artifacts/model/plots/roc-multiclass.html',\n",
       "   'model': 'store://artifacts/function-marketplace/sklearn_linear_model_LogisticRegression_model:0326c1bd8f3d4f7a95719e6e51ff1483'}},\n",
       " {'sklearn_ensemble_AdaBoostClassifier': {'accuracy': 0.90625,\n",
       "   'test-error': 0.09375,\n",
       "   'auc-micro': 0.96044921875,\n",
       "   'auc-weighted': 1.0,\n",
       "   'f1-score': 0.886039886039886,\n",
       "   'precision_score': 0.875,\n",
       "   'recall_score': 0.9333333333333332,\n",
       "   'test_set': 'store://artifacts/function-marketplace/sklearn_ensemble_AdaBoostClassifier_test_set:d5386ba6222e41f8a9f46d96ba866bc0',\n",
       "   'confusion-matrix': 'v3io:///projects/function-marketplace/artifacts/model/plots/confusion-matrix.html',\n",
       "   'feature-importances': 'v3io:///projects/function-marketplace/artifacts/model/plots/feature-importances.html',\n",
       "   'precision-recall-multiclass': 'v3io:///projects/function-marketplace/artifacts/model/plots/precision-recall-multiclass.html',\n",
       "   'roc-multiclass': 'v3io:///projects/function-marketplace/artifacts/model/plots/roc-multiclass.html',\n",
       "   'model': 'store://artifacts/function-marketplace/sklearn_ensemble_AdaBoostClassifier_model:d5386ba6222e41f8a9f46d96ba866bc0'}}]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "outputs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Set model parameters and run remotely**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "> 2021-10-17 11:44:48,897 [info] starting run sklearn_ensemble_RandomForestClassifier uid=753aec85957f402587d562cd8a61be80 DB=http://mlrun-api:8080\n",
      "> 2021-10-17 11:44:49,037 [info] Job is running in the background, pod: sklearn-ensemble-randomforestclassifier-wqkk6\n",
      "> 2021-10-17 11:44:58,008 [info] run executed, status=completed\n",
      "final state: completed\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>\n",
       ".dictlist {\n",
       "  background-color: #4EC64B;\n",
       "  text-align: center;\n",
       "  margin: 4px;\n",
       "  border-radius: 3px; padding: 0px 3px 1px 3px; display: inline-block;}\n",
       ".artifact {\n",
       "  cursor: pointer;\n",
       "  background-color: #4EC64B;\n",
       "  text-align: left;\n",
       "  margin: 4px; border-radius: 3px; padding: 0px 3px 1px 3px; display: inline-block;\n",
       "}\n",
       "div.block.hidden {\n",
       "  display: none;\n",
       "}\n",
       ".clickable {\n",
       "  cursor: pointer;\n",
       "}\n",
       ".ellipsis {\n",
       "  display: inline-block;\n",
       "  max-width: 60px;\n",
       "  white-space: nowrap;\n",
       "  overflow: hidden;\n",
       "  text-overflow: ellipsis;\n",
       "}\n",
       ".master-wrapper {\n",
       "  display: flex;\n",
       "  flex-flow: row nowrap;\n",
       "  justify-content: flex-start;\n",
       "  align-items: stretch;\n",
       "}\n",
       ".master-tbl {\n",
       "  flex: 3\n",
       "}\n",
       ".master-wrapper > div {\n",
       "  margin: 4px;\n",
       "  padding: 10px;\n",
       "}\n",
       "iframe.fileview {\n",
       "  border: 0 none;\n",
       "  height: 100%;\n",
       "  width: 100%;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       ".pane-header-title {\n",
       "  width: 80%;\n",
       "  font-weight: 500;\n",
       "}\n",
       ".pane-header {\n",
       "  line-height: 1;\n",
       "  background-color: #4EC64B;\n",
       "  padding: 3px;\n",
       "}\n",
       ".pane-header .close {\n",
       "  font-size: 20px;\n",
       "  font-weight: 700;\n",
       "  float: right;\n",
       "  margin-top: -5px;\n",
       "}\n",
       ".master-wrapper .right-pane {\n",
       "  border: 1px inset silver;\n",
       "  width: 40%;\n",
       "  min-height: 300px;\n",
       "  flex: 3\n",
       "  min-width: 500px;\n",
       "}\n",
       ".master-wrapper * {\n",
       "  box-sizing: border-box;\n",
       "}\n",
       "</style><script>\n",
       "function copyToClipboard(fld) {\n",
       "    if (document.queryCommandSupported && document.queryCommandSupported('copy')) {\n",
       "        var textarea = document.createElement('textarea');\n",
       "        textarea.textContent = fld.innerHTML;\n",
       "        textarea.style.position = 'fixed';\n",
       "        document.body.appendChild(textarea);\n",
       "        textarea.select();\n",
       "\n",
       "        try {\n",
       "            return document.execCommand('copy'); // Security exception may be thrown by some browsers.\n",
       "        } catch (ex) {\n",
       "\n",
       "        } finally {\n",
       "            document.body.removeChild(textarea);\n",
       "        }\n",
       "    }\n",
       "}\n",
       "function expandPanel(el) {\n",
       "  const panelName = \"#\" + el.getAttribute('paneName');\n",
       "  console.log(el.title);\n",
       "\n",
       "  document.querySelector(panelName + \"-title\").innerHTML = el.title\n",
       "  iframe = document.querySelector(panelName + \"-body\");\n",
       "\n",
       "  const tblcss = `<style> body { font-family: Arial, Helvetica, sans-serif;}\n",
       "    #csv { margin-bottom: 15px; }\n",
       "    #csv table { border-collapse: collapse;}\n",
       "    #csv table td { padding: 4px 8px; border: 1px solid silver;} </style>`;\n",
       "\n",
       "  function csvToHtmlTable(str) {\n",
       "    return '<div id=\"csv\"><table><tr><td>' +  str.replace(/[\\n\\r]+$/g, '').replace(/[\\n\\r]+/g, '</td></tr><tr><td>')\n",
       "      .replace(/,/g, '</td><td>') + '</td></tr></table></div>';\n",
       "  }\n",
       "\n",
       "  function reqListener () {\n",
       "    if (el.title.endsWith(\".csv\")) {\n",
       "      iframe.setAttribute(\"srcdoc\", tblcss + csvToHtmlTable(this.responseText));\n",
       "    } else {\n",
       "      iframe.setAttribute(\"srcdoc\", this.responseText);\n",
       "    }\n",
       "    console.log(this.responseText);\n",
       "  }\n",
       "\n",
       "  const oReq = new XMLHttpRequest();\n",
       "  oReq.addEventListener(\"load\", reqListener);\n",
       "  oReq.open(\"GET\", el.title);\n",
       "  oReq.send();\n",
       "\n",
       "\n",
       "  //iframe.src = el.title;\n",
       "  const resultPane = document.querySelector(panelName + \"-pane\");\n",
       "  if (resultPane.classList.contains(\"hidden\")) {\n",
       "    resultPane.classList.remove(\"hidden\");\n",
       "  }\n",
       "}\n",
       "function closePanel(el) {\n",
       "  const panelName = \"#\" + el.getAttribute('paneName')\n",
       "  const resultPane = document.querySelector(panelName + \"-pane\");\n",
       "  if (!resultPane.classList.contains(\"hidden\")) {\n",
       "    resultPane.classList.add(\"hidden\");\n",
       "  }\n",
       "}\n",
       "\n",
       "</script>\n",
       "<div class=\"master-wrapper\">\n",
       "  <div class=\"block master-tbl\"><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>project</th>\n",
       "      <th>uid</th>\n",
       "      <th>iter</th>\n",
       "      <th>start</th>\n",
       "      <th>state</th>\n",
       "      <th>name</th>\n",
       "      <th>labels</th>\n",
       "      <th>inputs</th>\n",
       "      <th>parameters</th>\n",
       "      <th>results</th>\n",
       "      <th>artifacts</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>function-marketplace</td>\n",
       "      <td><div title=\"753aec85957f402587d562cd8a61be80\"><a href=\"https://dashboard.default-tenant.app.dev39.lab.iguazeng.com/mlprojects/function-marketplace/jobs/monitor/753aec85957f402587d562cd8a61be80/overview\" target=\"_blank\" >...8a61be80</a></div></td>\n",
       "      <td>0</td>\n",
       "      <td>Oct 17 11:44:55</td>\n",
       "      <td>completed</td>\n",
       "      <td>sklearn_ensemble_RandomForestClassifier</td>\n",
       "      <td><div class=\"dictlist\">v3io_user=dani</div><div class=\"dictlist\">kind=job</div><div class=\"dictlist\">owner=dani</div><div class=\"dictlist\">host=sklearn-ensemble-randomforestclassifier-wqkk6</div><div class=\"dictlist\">class=sklearn.ensemble.RandomForestClassifier</div></td>\n",
       "      <td><div title=\"https://s3.wasabisys.com/iguazio/data/iris/iris_dataset.csv\">dataset</div></td>\n",
       "      <td><div class=\"dictlist\">model_pkg_class=sklearn.ensemble.RandomForestClassifier</div><div class=\"dictlist\">label_column=label</div><div class=\"dictlist\">CLASS_max_depth=5</div></td>\n",
       "      <td><div class=\"dictlist\">accuracy=0.90625</div><div class=\"dictlist\">test-error=0.09375</div><div class=\"dictlist\">auc-micro=0.99267578125</div><div class=\"dictlist\">auc-weighted=1.0</div><div class=\"dictlist\">f1-score=0.886039886039886</div><div class=\"dictlist\">precision_score=0.875</div><div class=\"dictlist\">recall_score=0.9333333333333332</div></td>\n",
       "      <td><div title=\"v3io:///projects/function-marketplace/artifacts/data/test_set.parquet\">test_set</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"resultfa48eb49\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/confusion-matrix.html\">confusion-matrix</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"resultfa48eb49\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/feature-importances.html\">feature-importances</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"resultfa48eb49\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/precision-recall-multiclass.html\">precision-recall-multiclass</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"resultfa48eb49\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/roc-multiclass.html\">roc-multiclass</div><div title=\"v3io:///projects/function-marketplace/artifacts/model/\">model</div></td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div></div>\n",
       "  <div id=\"resultfa48eb49-pane\" class=\"right-pane block hidden\">\n",
       "    <div class=\"pane-header\">\n",
       "      <span id=\"resultfa48eb49-title\" class=\"pane-header-title\">Title</span>\n",
       "      <span onclick=\"closePanel(this)\" paneName=\"resultfa48eb49\" class=\"close clickable\">&times;</span>\n",
       "    </div>\n",
       "    <iframe class=\"fileview\" id=\"resultfa48eb49-body\"></iframe>\n",
       "  </div>\n",
       "</div>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<b> > to track results use the .show() or .logs() methods  or <a href=\"https://dashboard.default-tenant.app.dev39.lab.iguazeng.com/mlprojects/function-marketplace/jobs/monitor/753aec85957f402587d562cd8a61be80/overview\" target=\"_blank\">click here</a> to open in UI</b>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "> 2021-10-17 11:44:58,256 [info] run executed, status=completed\n",
      "> 2021-10-17 11:44:58,257 [info] starting run sklearn_linear_model_LogisticRegression uid=c5714a4e2acb4b2e8c48a2ef77d9a05b DB=http://mlrun-api:8080\n",
      "> 2021-10-17 11:44:58,607 [info] Job is running in the background, pod: sklearn-linear-model-logisticregression-zxfgw\n",
      "> 2021-10-17 11:45:06,599 [info] run executed, status=completed\n",
      "final state: completed\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>\n",
       ".dictlist {\n",
       "  background-color: #4EC64B;\n",
       "  text-align: center;\n",
       "  margin: 4px;\n",
       "  border-radius: 3px; padding: 0px 3px 1px 3px; display: inline-block;}\n",
       ".artifact {\n",
       "  cursor: pointer;\n",
       "  background-color: #4EC64B;\n",
       "  text-align: left;\n",
       "  margin: 4px; border-radius: 3px; padding: 0px 3px 1px 3px; display: inline-block;\n",
       "}\n",
       "div.block.hidden {\n",
       "  display: none;\n",
       "}\n",
       ".clickable {\n",
       "  cursor: pointer;\n",
       "}\n",
       ".ellipsis {\n",
       "  display: inline-block;\n",
       "  max-width: 60px;\n",
       "  white-space: nowrap;\n",
       "  overflow: hidden;\n",
       "  text-overflow: ellipsis;\n",
       "}\n",
       ".master-wrapper {\n",
       "  display: flex;\n",
       "  flex-flow: row nowrap;\n",
       "  justify-content: flex-start;\n",
       "  align-items: stretch;\n",
       "}\n",
       ".master-tbl {\n",
       "  flex: 3\n",
       "}\n",
       ".master-wrapper > div {\n",
       "  margin: 4px;\n",
       "  padding: 10px;\n",
       "}\n",
       "iframe.fileview {\n",
       "  border: 0 none;\n",
       "  height: 100%;\n",
       "  width: 100%;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       ".pane-header-title {\n",
       "  width: 80%;\n",
       "  font-weight: 500;\n",
       "}\n",
       ".pane-header {\n",
       "  line-height: 1;\n",
       "  background-color: #4EC64B;\n",
       "  padding: 3px;\n",
       "}\n",
       ".pane-header .close {\n",
       "  font-size: 20px;\n",
       "  font-weight: 700;\n",
       "  float: right;\n",
       "  margin-top: -5px;\n",
       "}\n",
       ".master-wrapper .right-pane {\n",
       "  border: 1px inset silver;\n",
       "  width: 40%;\n",
       "  min-height: 300px;\n",
       "  flex: 3\n",
       "  min-width: 500px;\n",
       "}\n",
       ".master-wrapper * {\n",
       "  box-sizing: border-box;\n",
       "}\n",
       "</style><script>\n",
       "function copyToClipboard(fld) {\n",
       "    if (document.queryCommandSupported && document.queryCommandSupported('copy')) {\n",
       "        var textarea = document.createElement('textarea');\n",
       "        textarea.textContent = fld.innerHTML;\n",
       "        textarea.style.position = 'fixed';\n",
       "        document.body.appendChild(textarea);\n",
       "        textarea.select();\n",
       "\n",
       "        try {\n",
       "            return document.execCommand('copy'); // Security exception may be thrown by some browsers.\n",
       "        } catch (ex) {\n",
       "\n",
       "        } finally {\n",
       "            document.body.removeChild(textarea);\n",
       "        }\n",
       "    }\n",
       "}\n",
       "function expandPanel(el) {\n",
       "  const panelName = \"#\" + el.getAttribute('paneName');\n",
       "  console.log(el.title);\n",
       "\n",
       "  document.querySelector(panelName + \"-title\").innerHTML = el.title\n",
       "  iframe = document.querySelector(panelName + \"-body\");\n",
       "\n",
       "  const tblcss = `<style> body { font-family: Arial, Helvetica, sans-serif;}\n",
       "    #csv { margin-bottom: 15px; }\n",
       "    #csv table { border-collapse: collapse;}\n",
       "    #csv table td { padding: 4px 8px; border: 1px solid silver;} </style>`;\n",
       "\n",
       "  function csvToHtmlTable(str) {\n",
       "    return '<div id=\"csv\"><table><tr><td>' +  str.replace(/[\\n\\r]+$/g, '').replace(/[\\n\\r]+/g, '</td></tr><tr><td>')\n",
       "      .replace(/,/g, '</td><td>') + '</td></tr></table></div>';\n",
       "  }\n",
       "\n",
       "  function reqListener () {\n",
       "    if (el.title.endsWith(\".csv\")) {\n",
       "      iframe.setAttribute(\"srcdoc\", tblcss + csvToHtmlTable(this.responseText));\n",
       "    } else {\n",
       "      iframe.setAttribute(\"srcdoc\", this.responseText);\n",
       "    }\n",
       "    console.log(this.responseText);\n",
       "  }\n",
       "\n",
       "  const oReq = new XMLHttpRequest();\n",
       "  oReq.addEventListener(\"load\", reqListener);\n",
       "  oReq.open(\"GET\", el.title);\n",
       "  oReq.send();\n",
       "\n",
       "\n",
       "  //iframe.src = el.title;\n",
       "  const resultPane = document.querySelector(panelName + \"-pane\");\n",
       "  if (resultPane.classList.contains(\"hidden\")) {\n",
       "    resultPane.classList.remove(\"hidden\");\n",
       "  }\n",
       "}\n",
       "function closePanel(el) {\n",
       "  const panelName = \"#\" + el.getAttribute('paneName')\n",
       "  const resultPane = document.querySelector(panelName + \"-pane\");\n",
       "  if (!resultPane.classList.contains(\"hidden\")) {\n",
       "    resultPane.classList.add(\"hidden\");\n",
       "  }\n",
       "}\n",
       "\n",
       "</script>\n",
       "<div class=\"master-wrapper\">\n",
       "  <div class=\"block master-tbl\"><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>project</th>\n",
       "      <th>uid</th>\n",
       "      <th>iter</th>\n",
       "      <th>start</th>\n",
       "      <th>state</th>\n",
       "      <th>name</th>\n",
       "      <th>labels</th>\n",
       "      <th>inputs</th>\n",
       "      <th>parameters</th>\n",
       "      <th>results</th>\n",
       "      <th>artifacts</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>function-marketplace</td>\n",
       "      <td><div title=\"c5714a4e2acb4b2e8c48a2ef77d9a05b\"><a href=\"https://dashboard.default-tenant.app.dev39.lab.iguazeng.com/mlprojects/function-marketplace/jobs/monitor/c5714a4e2acb4b2e8c48a2ef77d9a05b/overview\" target=\"_blank\" >...77d9a05b</a></div></td>\n",
       "      <td>0</td>\n",
       "      <td>Oct 17 11:45:04</td>\n",
       "      <td>completed</td>\n",
       "      <td>sklearn_linear_model_LogisticRegression</td>\n",
       "      <td><div class=\"dictlist\">v3io_user=dani</div><div class=\"dictlist\">kind=job</div><div class=\"dictlist\">owner=dani</div><div class=\"dictlist\">host=sklearn-linear-model-logisticregression-zxfgw</div><div class=\"dictlist\">class=sklearn.linear_model.LogisticRegression</div></td>\n",
       "      <td><div title=\"https://s3.wasabisys.com/iguazio/data/iris/iris_dataset.csv\">dataset</div></td>\n",
       "      <td><div class=\"dictlist\">model_pkg_class=sklearn.linear_model.LogisticRegression</div><div class=\"dictlist\">label_column=label</div><div class=\"dictlist\">CLASS_solver=liblinear</div></td>\n",
       "      <td><div class=\"dictlist\">accuracy=0.875</div><div class=\"dictlist\">test-error=0.125</div><div class=\"dictlist\">auc-micro=0.97265625</div><div class=\"dictlist\">auc-weighted=0.997004357298475</div><div class=\"dictlist\">f1-score=0.8534798534798534</div><div class=\"dictlist\">precision_score=0.8518518518518517</div><div class=\"dictlist\">recall_score=0.9111111111111111</div></td>\n",
       "      <td><div title=\"v3io:///projects/function-marketplace/artifacts/data/test_set.parquet\">test_set</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"resultbef08456\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/confusion-matrix.html\">confusion-matrix</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"resultbef08456\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/precision-recall-multiclass.html\">precision-recall-multiclass</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"resultbef08456\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/roc-multiclass.html\">roc-multiclass</div><div title=\"v3io:///projects/function-marketplace/artifacts/model/\">model</div></td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div></div>\n",
       "  <div id=\"resultbef08456-pane\" class=\"right-pane block hidden\">\n",
       "    <div class=\"pane-header\">\n",
       "      <span id=\"resultbef08456-title\" class=\"pane-header-title\">Title</span>\n",
       "      <span onclick=\"closePanel(this)\" paneName=\"resultbef08456\" class=\"close clickable\">&times;</span>\n",
       "    </div>\n",
       "    <iframe class=\"fileview\" id=\"resultbef08456-body\"></iframe>\n",
       "  </div>\n",
       "</div>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<b> > to track results use the .show() or .logs() methods  or <a href=\"https://dashboard.default-tenant.app.dev39.lab.iguazeng.com/mlprojects/function-marketplace/jobs/monitor/c5714a4e2acb4b2e8c48a2ef77d9a05b/overview\" target=\"_blank\">click here</a> to open in UI</b>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "> 2021-10-17 11:45:07,795 [info] run executed, status=completed\n",
      "> 2021-10-17 11:45:07,797 [info] starting run sklearn_ensemble_AdaBoostClassifier uid=4b042898a9b94bb1bee72f788a163a10 DB=http://mlrun-api:8080\n",
      "> 2021-10-17 11:45:07,926 [info] Job is running in the background, pod: sklearn-ensemble-adaboostclassifier-nphgz\n",
      "> 2021-10-17 11:45:17,094 [info] run executed, status=completed\n",
      "final state: completed\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>\n",
       ".dictlist {\n",
       "  background-color: #4EC64B;\n",
       "  text-align: center;\n",
       "  margin: 4px;\n",
       "  border-radius: 3px; padding: 0px 3px 1px 3px; display: inline-block;}\n",
       ".artifact {\n",
       "  cursor: pointer;\n",
       "  background-color: #4EC64B;\n",
       "  text-align: left;\n",
       "  margin: 4px; border-radius: 3px; padding: 0px 3px 1px 3px; display: inline-block;\n",
       "}\n",
       "div.block.hidden {\n",
       "  display: none;\n",
       "}\n",
       ".clickable {\n",
       "  cursor: pointer;\n",
       "}\n",
       ".ellipsis {\n",
       "  display: inline-block;\n",
       "  max-width: 60px;\n",
       "  white-space: nowrap;\n",
       "  overflow: hidden;\n",
       "  text-overflow: ellipsis;\n",
       "}\n",
       ".master-wrapper {\n",
       "  display: flex;\n",
       "  flex-flow: row nowrap;\n",
       "  justify-content: flex-start;\n",
       "  align-items: stretch;\n",
       "}\n",
       ".master-tbl {\n",
       "  flex: 3\n",
       "}\n",
       ".master-wrapper > div {\n",
       "  margin: 4px;\n",
       "  padding: 10px;\n",
       "}\n",
       "iframe.fileview {\n",
       "  border: 0 none;\n",
       "  height: 100%;\n",
       "  width: 100%;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       ".pane-header-title {\n",
       "  width: 80%;\n",
       "  font-weight: 500;\n",
       "}\n",
       ".pane-header {\n",
       "  line-height: 1;\n",
       "  background-color: #4EC64B;\n",
       "  padding: 3px;\n",
       "}\n",
       ".pane-header .close {\n",
       "  font-size: 20px;\n",
       "  font-weight: 700;\n",
       "  float: right;\n",
       "  margin-top: -5px;\n",
       "}\n",
       ".master-wrapper .right-pane {\n",
       "  border: 1px inset silver;\n",
       "  width: 40%;\n",
       "  min-height: 300px;\n",
       "  flex: 3\n",
       "  min-width: 500px;\n",
       "}\n",
       ".master-wrapper * {\n",
       "  box-sizing: border-box;\n",
       "}\n",
       "</style><script>\n",
       "function copyToClipboard(fld) {\n",
       "    if (document.queryCommandSupported && document.queryCommandSupported('copy')) {\n",
       "        var textarea = document.createElement('textarea');\n",
       "        textarea.textContent = fld.innerHTML;\n",
       "        textarea.style.position = 'fixed';\n",
       "        document.body.appendChild(textarea);\n",
       "        textarea.select();\n",
       "\n",
       "        try {\n",
       "            return document.execCommand('copy'); // Security exception may be thrown by some browsers.\n",
       "        } catch (ex) {\n",
       "\n",
       "        } finally {\n",
       "            document.body.removeChild(textarea);\n",
       "        }\n",
       "    }\n",
       "}\n",
       "function expandPanel(el) {\n",
       "  const panelName = \"#\" + el.getAttribute('paneName');\n",
       "  console.log(el.title);\n",
       "\n",
       "  document.querySelector(panelName + \"-title\").innerHTML = el.title\n",
       "  iframe = document.querySelector(panelName + \"-body\");\n",
       "\n",
       "  const tblcss = `<style> body { font-family: Arial, Helvetica, sans-serif;}\n",
       "    #csv { margin-bottom: 15px; }\n",
       "    #csv table { border-collapse: collapse;}\n",
       "    #csv table td { padding: 4px 8px; border: 1px solid silver;} </style>`;\n",
       "\n",
       "  function csvToHtmlTable(str) {\n",
       "    return '<div id=\"csv\"><table><tr><td>' +  str.replace(/[\\n\\r]+$/g, '').replace(/[\\n\\r]+/g, '</td></tr><tr><td>')\n",
       "      .replace(/,/g, '</td><td>') + '</td></tr></table></div>';\n",
       "  }\n",
       "\n",
       "  function reqListener () {\n",
       "    if (el.title.endsWith(\".csv\")) {\n",
       "      iframe.setAttribute(\"srcdoc\", tblcss + csvToHtmlTable(this.responseText));\n",
       "    } else {\n",
       "      iframe.setAttribute(\"srcdoc\", this.responseText);\n",
       "    }\n",
       "    console.log(this.responseText);\n",
       "  }\n",
       "\n",
       "  const oReq = new XMLHttpRequest();\n",
       "  oReq.addEventListener(\"load\", reqListener);\n",
       "  oReq.open(\"GET\", el.title);\n",
       "  oReq.send();\n",
       "\n",
       "\n",
       "  //iframe.src = el.title;\n",
       "  const resultPane = document.querySelector(panelName + \"-pane\");\n",
       "  if (resultPane.classList.contains(\"hidden\")) {\n",
       "    resultPane.classList.remove(\"hidden\");\n",
       "  }\n",
       "}\n",
       "function closePanel(el) {\n",
       "  const panelName = \"#\" + el.getAttribute('paneName')\n",
       "  const resultPane = document.querySelector(panelName + \"-pane\");\n",
       "  if (!resultPane.classList.contains(\"hidden\")) {\n",
       "    resultPane.classList.add(\"hidden\");\n",
       "  }\n",
       "}\n",
       "\n",
       "</script>\n",
       "<div class=\"master-wrapper\">\n",
       "  <div class=\"block master-tbl\"><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>project</th>\n",
       "      <th>uid</th>\n",
       "      <th>iter</th>\n",
       "      <th>start</th>\n",
       "      <th>state</th>\n",
       "      <th>name</th>\n",
       "      <th>labels</th>\n",
       "      <th>inputs</th>\n",
       "      <th>parameters</th>\n",
       "      <th>results</th>\n",
       "      <th>artifacts</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>function-marketplace</td>\n",
       "      <td><div title=\"4b042898a9b94bb1bee72f788a163a10\"><a href=\"https://dashboard.default-tenant.app.dev39.lab.iguazeng.com/mlprojects/function-marketplace/jobs/monitor/4b042898a9b94bb1bee72f788a163a10/overview\" target=\"_blank\" >...8a163a10</a></div></td>\n",
       "      <td>0</td>\n",
       "      <td>Oct 17 11:45:14</td>\n",
       "      <td>completed</td>\n",
       "      <td>sklearn_ensemble_AdaBoostClassifier</td>\n",
       "      <td><div class=\"dictlist\">v3io_user=dani</div><div class=\"dictlist\">kind=job</div><div class=\"dictlist\">owner=dani</div><div class=\"dictlist\">host=sklearn-ensemble-adaboostclassifier-nphgz</div><div class=\"dictlist\">class=sklearn.ensemble.AdaBoostClassifier</div></td>\n",
       "      <td><div title=\"https://s3.wasabisys.com/iguazio/data/iris/iris_dataset.csv\">dataset</div></td>\n",
       "      <td><div class=\"dictlist\">model_pkg_class=sklearn.ensemble.AdaBoostClassifier</div><div class=\"dictlist\">label_column=label</div><div class=\"dictlist\">CLASS_n_estimators=200</div><div class=\"dictlist\">CLASS_learning_rate=0.01</div></td>\n",
       "      <td><div class=\"dictlist\">accuracy=0.90625</div><div class=\"dictlist\">test-error=0.09375</div><div class=\"dictlist\">auc-micro=0.96044921875</div><div class=\"dictlist\">auc-weighted=1.0</div><div class=\"dictlist\">f1-score=0.886039886039886</div><div class=\"dictlist\">precision_score=0.875</div><div class=\"dictlist\">recall_score=0.9333333333333332</div></td>\n",
       "      <td><div title=\"v3io:///projects/function-marketplace/artifacts/data/test_set.parquet\">test_set</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result86f4f28f\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/confusion-matrix.html\">confusion-matrix</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result86f4f28f\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/feature-importances.html\">feature-importances</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result86f4f28f\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/precision-recall-multiclass.html\">precision-recall-multiclass</div><div class=\"artifact\" onclick=\"expandPanel(this)\" paneName=\"result86f4f28f\" title=\"files/v3io/projects/function-marketplace/artifacts/model/plots/roc-multiclass.html\">roc-multiclass</div><div title=\"v3io:///projects/function-marketplace/artifacts/model/\">model</div></td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div></div>\n",
       "  <div id=\"result86f4f28f-pane\" class=\"right-pane block hidden\">\n",
       "    <div class=\"pane-header\">\n",
       "      <span id=\"result86f4f28f-title\" class=\"pane-header-title\">Title</span>\n",
       "      <span onclick=\"closePanel(this)\" paneName=\"result86f4f28f\" class=\"close clickable\">&times;</span>\n",
       "    </div>\n",
       "    <iframe class=\"fileview\" id=\"result86f4f28f-body\"></iframe>\n",
       "  </div>\n",
       "</div>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<b> > to track results use the .show() or .logs() methods  or <a href=\"https://dashboard.default-tenant.app.dev39.lab.iguazeng.com/mlprojects/function-marketplace/jobs/monitor/4b042898a9b94bb1bee72f788a163a10/overview\" target=\"_blank\">click here</a> to open in UI</b>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "> 2021-10-17 11:45:27,240 [info] run executed, status=completed\n"
     ]
    }
   ],
   "source": [
    "outputs = []\n",
    "for model in models:\n",
    "    task_copy = task_params.copy()\n",
    "    task_copy.update(\n",
    "        {\n",
    "            \"params\":{ \"model_pkg_class\" : model,\n",
    "                       \"label_column\"    : \"label\"}\n",
    "        }\n",
    "    )\n",
    "    \n",
    "    # customize specific model parameters\n",
    "    if \"RandomForestClassifier\" in model:\n",
    "        task_copy[\"params\"].update({\"CLASS_max_depth\" : 5})\n",
    "\n",
    "    if \"LogisticRegression\" in model:\n",
    "        task_copy[\"params\"].update({\"CLASS_solver\" : \"liblinear\"})\n",
    "    \n",
    "    if \"AdaBoostClassifier\" in model:\n",
    "        task_copy[\"params\"].update({\"CLASS_n_estimators\"  : 200,\n",
    "                                    \"CLASS_learning_rate\" : 0.01\n",
    "                                   })\n",
    "    \n",
    "    name = model.replace('.', '_')\n",
    "    output = fn.run(mlrun.NewTask(**task_copy),\n",
    "                    name=name,\n",
    "                    inputs={\"dataset\"  : data_url},\n",
    "                    local=False)\n",
    "    outputs.append({name: output.outputs})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Back to the top](#Generic-scikit-learn-classifier)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:root] *",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}