011_classification.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from util import render_audio_sample"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Digging Into Deep Time and Deep Cover\n",
    "\n",
    "<a href=\"https://doi.org/10.5281/zenodo.3875779\"><img src=\"https://zenodo.org/badge/DOI/10.5281/zenodo.3875779.svg\" align=\"right\" alt=\"doi: 10.5281/zenodo.3875779\" style=\"padding: 0px 10px 10px 0px\"></a>\n",
    "<a href=\"https://github.com/morganjwilliams/gs2020-diggingdeeper/blob/master/LICENSE\"><img src=\"https://img.shields.io/badge/License-MIT-blue.svg\" align=\"right\" alt=\"License: MIT\" style=\"padding: 0px 10px 10px 0px\"></a>\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "<span id='authors'><b>Morgan Williams <a class=\"fa fa-twitter\" aria-hidden=\"true\" href=\"https://twitter.com/metasomite\" title=\"@metasomite\"></a></b>, Jens Klump, Steve Barnes and Fang Huang; </span>\n",
    "<span id='affiliation'><em>CSIRO Mineral Resources</em></span>\n",
    "\n",
    "\n",
    "### Contents\n",
    "\n",
    "| [**Abstract**](./00_overview.ipynb) | **Introduction**                                                    | [**Examples**](./00_overview.ipynb#Examples)            | [**Tools**](./00_overview.ipynb#Leveraging-the-Scientific-Python-Ecosystem) | \n",
    "|:-----|:-----|:-----|:-----|\n",
    "|  | [Minerals Exploration](./00_overview.ipynb#An-Evolving-Role-of-Geochemistry-in-Mineral-Exploration)  | [Classification](./011_classification.ipynb) |  |\n",
    "|  | [Data Driven Geochem](./00_overview.ipynb#Data-Driven-Geochemistry) | [Data Visualization](./012_datavis.ipynb) |  |"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Classification: Towards Multivariate Classifiers\n",
    "\n",
    "In geochemistry, classification is for historical reasons often treated as a graphical or two-dimensional problem for which the output includes a graphically-represented classification or discrimination diagram (driven by a need to visualize the data; e.g. [Pearce, 1973]). And while many of these diagrams have been practically useful for 'binning' and describing rocks, the divisions we use don't necessarily make sense in the natural world (e.g. consider the Total Alkali - Silica diagram). Modern statistical and machine learning techniques can readily be applied to make better use of the data dimensionality and move beyond the human-centric limitations. In this case a data driven approach instead considers divisions according to the natural clustering of data, where they exist.\n",
    "\n",
    "**Key Messages**:\n",
    "* Use the dimensionality you have\n",
    "* Particularly if you're using major element data, use compositional transforms\n",
    "* Where available, use metrics to understand estimated uncertainty and confidence\n",
    "* Try to visualize and understand what's happening in lower dimensional spaces\n",
    "\n",
    "### Searching for Enriched Basalts\n",
    "\n",
    "This example is motivated by one of the research questions for my postdoc: where and when do we find the earliest examples of enriched basalts (analogous to modern day OIB)? Large magmatic nickel deposits are assocaited with enriched basaltic suites, and if we can generate spatio-temporal constraints on their genesis (particularly in geological time), we might be able to reduce the nickel exploration search space considerably. \n",
    "\n",
    "One approach to this question is to formulate it as a classification problem. Whole rock geochemistry is widely used to constrain the tectonic context in which magmatic rocks have been formed, with varying levels of success. Tectonic discrimination using whole-rock geochemical data is complicated by long geological histories, alteration, biased preservation and secular change. Notably, there is a lack of confidently identified oceanic crust in the Archean, suggesting finding OIB-like rocks is likely even more challenging. Most traditional tectonic discrimination diagrams use geochemical proxies to encode information about geological processes or (e.g. the \"Pearce diagrams\" which use Th/Yb, Nb/Yb and TiO2/Yb as proxies for subduction influence, source enrichment and depth of mantle melting respectively; [Pearce, 2008]), while others are designed to provide optimal discrimination for a specific scenario. While a wide variety of geochemical proxies can be used with relative confidence to discriminate between modern rocks, investigating primary geochemical contrast in ancient rocks (especially into the Archean) is typically considerd to require restriction of variables considered to those which are least effected by geological modification, i.e. immobile elements such as the High Field Strength Elements (HFSE) and Rare Earth Elements (REE). Overall the limited perspective of graphically-focused methods can lead to variable and uncertain classifications depending on the elements chosen, especially when applied into deep time. \n",
    "\n",
    "The example below goes through some of the first steps of exploring how we might be able to classify Paleoproteroizic and Archean basalts based on their geochemistry with a machine learning approach which capitalizes on the dimensionality of our datasets. But before we start digging into deep time (also see the [note on dealing with space and time](./00_overview.ipynb#Other-Opportunities-in-Data-Integration-and-Exploration) in the overview notebook), we first consider how well we can distinguish modern ocean island basalts. More so than the results of the prediction, we'd like to demonstrate and explore part of the process.\n",
    "\n",
    "### Machine Tectonic Discrimination\n",
    "\n",
    "Below we'll demonstrate a machine learning approach to classifying basalt geochemistry based on tectonic setting. We'll adapt a version of a support vector classifier, similar to those previously used for this problem ([Petrelli, 2016]; [Ueki, 2018]), and re-use the same 29-dimension tectonic discrimination dataset (just over 2000 samples taken from global repositories, each with majors, minors, traces and isotopes) from Ueki et al. (2018), modified to record the tectonic setting of each basalt in a 'Class' column.\n",
    "\n",
    "First let's examine a well-used trace element discrimination diagram (the Th-Nb-Yb diagram; [Pearce, 2008]) to get some insight what some of the key issues with classification using only a few elements or dimension. This is one of number of classification and tectonic diagrams have been developed which use specific geochemical proxies to link rock and mineral geochemistry to geological processes and reservoirs (here Nb/Yb and Th/Yb which distinguish relative enrichment of the mantle and arc-related processes). As the diagram presents only two dimensions (or three, depending who you talk to), the data presented show only part of the whole picture. This reduced dimensionality means that when you compare rocks from a variety of settings, you're likely to see large degrees of overlap, as found by [Li et al. (2015)], and shown below. While the diagram is not particularly useful as a classification tool, the proxies used (ratios or otherwise) still provide useful 'features' which can be added to datasets before constructing a multivariate classifier.\n",
    "\n",
    "| <img src=\"https://github.com/morganjwilliams/gs2020-diggingdeeper/raw/develop/img/Smithies2018Fig1.png\" style=\"display:inline; margin: 0px 15px 15px 0px;\" width=\"70%\"/>|<img src=\"https://github.com/morganjwilliams/gs2020-diggingdeeper/raw/develop/img/Li2015Fig9.png\" style=\"display:inline; margin: 0px 15px 15px 0px;\" width=\"70%\"/> |\n",
    "|--|--|\n",
    "| Figure from [Smithies (2018)] illustrating the main features of the Th-Yb-Nb diagram [Pearce (2008)], including the mantle or 'MORB-OIB array' and the 'modern arc array'. | Figure from [Li et al. (2015)], illustrating the significant degree of overlap in these two dimensions between different tectonic settings renders this approach futile for generalized discrimination.  |\n",
    "\n",
    "Below we've loaded the dataset from [Ueki, 2018] and added the Nb/Yb and Th/Yb, and contoured data density for basalts from three different tectonic settings (at 30th, 50th and 80th percentiles). Notably, these data density contours largely line up with what we expect from the Pearce diagram.\n",
    "\n",
    "<!--(BAB: Back Arc Basin, CA: Continental Arc, CFB: Continental Flood Basalt, IA: Island Arc, IOA: Intra-oceanic Arc, MOR: Mid-Ocean Ridge, OI: Ocean Island, OP: Oceanic Plateau) -->\n",
    "\n",
    "[Pearce, 1973]: https://doi.org/10.1016/0012-821X(73)90129-5 \"Pearce, J.A., Cann, J.R., 1973. Tectonic setting of basic volcanic rocks determined using trace element analyses. Earth and Planetary Science Letters 19, 290–300. doi: 10.1016/0012-821X(73)90129-5\"\n",
    "\n",
    "[Li et al. (2015)]: https://doi.org/10.1016/j.lithos.2015.06.022 \"Li, C., Arndt, N.T., Tang, Q., Ripley, E.M., 2015. Trace element indiscrimination diagrams. Lithos 232, 76–83. \"\n",
    "\n",
    "[Smithies (2018)]: https://doi.org/10.1016/j.epsl.2018.01.034 \"Smithies, R.H., Ivanic, T.J., Lowrey, J.R., Morris, P.A., Barnes, S.J., Wyche, S., Lu, Y.-J., 2018. Two distinct origins for Archean greenstone belts. Earth and Planetary Science Letters 487, 106–116.\"\n",
    "\n",
    "[Pearce (2008)]: https://doi.org/10.1016/j.lithos.2007.06.016 \"Pearce, J.A., 2008. Geochemical fingerprinting of oceanic basalts with applications to ophiolite classification and the search for Archean oceanic crust. Lithos 100, 14–48.\"\n",
    "\n",
    "[Pearce, 2008]: https://doi.org/10.1016/j.lithos.2007.06.016 \"Pearce, J.A., 2008. Geochemical fingerprinting of oceanic basalts with applications to ophiolite classification and the search for Archean oceanic crust. Lithos 100, 14–48.\"\n",
    "\n",
    "[Petrelli, 2016]: https://doi.org/10.1007/s00410-016-1292-2 \"Petrelli, M., Perugini, D., 2016. Solving petrological problems through machine learning: the study case of tectonic discrimination using geochemical and isotopic data. Contrib Mineral Petrol 171, 81.\"\n",
    "\n",
    "[Ueki, 2018]: https://doi.org/10.1029/2017GC007401 \"Ueki, K., Hino, H., Kuwatani, T., 2018. Geochemical Discrimination and Characteristics of Magmatic Tectonic Settings: A Machine-Learning-Based Approach. Geochemistry, Geophysics, Geosystems 19, 1327–1347.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import pyrolite.geochem\n",
    "\n",
    "df = pd.read_csv(\"https://storage.googleapis.com/aegc2019/ueki2018.csv\")\n",
    "df.pyrochem.add_ratio(\"Th/Yb\")\n",
    "df.pyrochem.add_ratio(\"Nb/Yb\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pyrolite.plot\n",
    "import matplotlib.pyplot as plt\n",
    "from pyrolite.plot.templates.pearce import pearceThNbYb\n",
    "from pyrolite.util.plot.legend import proxy_line\n",
    "\n",
    "fig, ax = plt.subplots(1, figsize=(8, 6))\n",
    "pearceThNbYb(ax)\n",
    "\n",
    "proxies = []\n",
    "labels = []\n",
    "for setting, ID, cmap in [(\"MORB\", \"MOR\", \"Blues\"),\n",
    "                          (\"Continental Arcs\", \"CA\",\"Reds\"),\n",
    "                          (\"Ocean Islands\", \"OI\", \"Greens\")]:\n",
    "    df.loc[(df.Class == ID), [\"Nb/Yb\", \"Th/Yb\"]].pyroplot.density(\n",
    "        ax=ax, logx=True, logy=True, bins=100, contours=[0.3, 0.5, 0.8], cmap=cmap\n",
    "    )\n",
    "    labels.append(setting)\n",
    "    proxies.append(proxy_line(color=plt.get_cmap(cmap)(0.5)))\n",
    "\n",
    "ax.legend(proxies, \n",
    "          labels,\n",
    "          facecolor=None,\n",
    "          frameon=False, \n",
    "          bbox_to_anchor=(1,0), \n",
    "          loc='lower right', \n",
    "          fontsize=12);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "However, when we add other tectonic settings to the diagram, we find something reminiscent of Li's diagram shown above, with a large degree of overlap:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "for setting, ID, cmap in [(\"Back Arcs\", \"BAB\", \"Purples\")]:\n",
    "    df.loc[(df.Class == ID), [\"Nb/Yb\", \"Th/Yb\"]].pyroplot.density(\n",
    "        ax=ax, logx=True, logy=True, bins=100, contours=[0.3, 0.5, 0.8], cmap=cmap\n",
    "    )\n",
    "    labels.append(setting)\n",
    "    proxies.append(proxy_line(color=plt.get_cmap(cmap)(0.5)))\n",
    "    \n",
    "ax.legend(proxies, \n",
    "          labels,\n",
    "          facecolor=None,\n",
    "          frameon=False, \n",
    "          bbox_to_anchor=(1,0), \n",
    "          loc='lower right', \n",
    "          fontsize=12)\n",
    "ax.figure"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Building and Testing Models\n",
    "\n",
    "Below we'll build a machine tectonic discrimination 'pipeline', including necessary data pre-processing, training and selecting a classifier, and assessing the overall performance of this classification model.\n",
    "\n",
    "\n",
    "\n",
    "There are a variety of types of classifiers we could use to achieve this, and here we've just picked one to work with which is relatively robust and can handle the dimensionality - a support vector classifier (SVC) with a radial basis function. We've included an illustrative example below for a dataset with two different groups of samples (two 'classes'), highlighting some of the differences between some commonly used classifiers.\n",
    "\n",
    "| <img src=\"https://github.com/morganjwilliams/gs2020-diggingdeeper/raw/develop/img/ClassifierComparison.png\" style=\"display:inline; margin: 0px 15px 15px 0px;\" width=\"50%\"/>  |\n",
    "|:--:|\n",
    "| Examples of decision surfaces of different classifier models for the same two-class dataset.|\n",
    "\n",
    "#### Pre-Processing\n",
    "\n",
    "To make the most of this dataset and build a more precise classifier, we need to make the most of the dataset dimensionality, as using more dimensions gives us more discriminating power. Making a machine learning classifier with these same two dimensions won't get us any further!\n",
    "\n",
    "Most statistical and machine learning algorithms perform best on scaled and approximately normally-distributed data. As we're using major elements to build our classification model, we should **consider using compositional log-ratio transforms** ([Aitchison, 1984]) before we train our model, such that spurious correlation due to closure is minimized (note that we're *not* transforming the isotope ratios here). Also, we should **include a scaling step** to transform our input values such that they fall within $[-1, 1]$, such that magnitude differences between variables are accounted for.\n",
    "\n",
    "[Aitchison, 1984]: https://doi.org/10.1007/BF01029316 \"Aitchison, J., 1984. The statistical analysis of geochemical compositions. Journal of the International Association for Mathematical Geology 16, 531–564. doi: 10.1007/BF01029316\"\n",
    "\n",
    "#### Training a Model\n",
    "\n",
    "Within our pipeline, we've specified a range of parameters for the classifier model ('hyperparameters') which it iterates through over a grid to find the most performant model. To ensure that our classifier is robust, we set aside independent subsets of our data for training and testing (used to fit and asses the model, respectively - typically about 30% for testing). We also use cross validation and ensure that separate subsets of the training data are used for fitting and validating models during model selection and tuning processes. Specifically here we used 5-fold stratified cross-validation, which splits the data into five separate representative subsets, and iterates through using each as the validation set. \n",
    "\n",
    "After this process has completed, the held-back test set can be used to independently verify the performance of the classifier model(s). If working as expected and with the desired accuracy and precision, we can extract the best classifier and put it to work on a new dataset! Below we train a classifier using our dataset, print out a quick diagrammatic summary (click for more details) and dig into the overall performance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import joblib\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import set_config\n",
    "from sklearn.pipeline import FeatureUnion, make_pipeline\n",
    "set_config(display='diagram')\n",
    "\n",
    "from pyrolite.util.skl.select import CompositionalSelector\n",
    "from pyrolite.util.skl.vis import plot_confusion_matrix\n",
    "from pyrolite.util.skl.pipeline import SVC_pipeline\n",
    "from pyrolite.util.skl.transform import CLRTransform, LogTransform "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>div.sk-top-container {color: black;background-color: white;}div.sk-toggleable {background-color: white;}label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.2em 0.3em;box-sizing: border-box;text-align: center;}div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}div.sk-estimator {font-family: monospace;background-color: #f0f8ff;margin: 0.25em 0.25em;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;}div.sk-estimator:hover {background-color: #d4ebff;}div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;}div.sk-item {z-index: 1;}div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}div.sk-parallel-item:only-child::after {width: 0;}div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0.2em;box-sizing: border-box;padding-bottom: 0.1em;background-color: white;position: relative;}div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}div.sk-label-container {position: relative;z-index: 2;text-align: center;}div.sk-container {display: inline-block;position: relative;}</style><div class=\"sk-top-container\"><div class=\"sk-container\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"69f922d9-93dc-4238-826e-2e8b2c481780\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"69f922d9-93dc-4238-826e-2e8b2c481780\">GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(cv=5,\n",
       "             estimator=Pipeline(steps=[('featureunion',\n",
       "                                        FeatureUnion(transformer_list=[('abundances '\n",
       "                                                                        '--> '\n",
       "                                                                        'CLR',\n",
       "                                                                        Pipeline(steps=[('compositionalselector',\n",
       "                                                                                         CompositionalSelector()),\n",
       "                                                                                        ('clrtransform',\n",
       "                                                                                         CLRTransform())])),\n",
       "                                                                       ('isotopes',\n",
       "                                                                        Pipeline(steps=[('compositionalselector',\n",
       "                                                                                         CompositionalSelector(inverse=True)),\n",
       "                                                                                        ('logtransform',\n",
       "                                                                                         LogTransform())]))])),\n",
       "                                       ('standardscaler', StandardScaler()),\n",
       "                                       ('svc',\n",
       "                                        SVC(cache_size=500,\n",
       "                                            class_weight='balanced',\n",
       "                                            decision_function_shape='ovo',\n",
       "                                            probability=True))]),\n",
       "             n_jobs=1,\n",
       "             param_grid={'svc__C': array([  0.31622777,   1.33352143,   5.62341325,  23.71373706,\n",
       "       100.        ]),\n",
       "                         'svc__gamma': array([0.00501187, 0.01154782, 0.02660725, 0.06130558, 0.14125375])})</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"7e20d7bc-2fd7-4447-9cf4-3e4f3476baf3\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"7e20d7bc-2fd7-4447-9cf4-3e4f3476baf3\">featureunion: FeatureUnion</label><div class=\"sk-toggleable__content\"><pre>FeatureUnion(transformer_list=[('abundances --> CLR',\n",
       "                                Pipeline(steps=[('compositionalselector',\n",
       "                                                 CompositionalSelector()),\n",
       "                                                ('clrtransform',\n",
       "                                                 CLRTransform())])),\n",
       "                               ('isotopes',\n",
       "                                Pipeline(steps=[('compositionalselector',\n",
       "                                                 CompositionalSelector(inverse=True)),\n",
       "                                                ('logtransform',\n",
       "                                                 LogTransform())]))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>abundances --&gt; CLR</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"0db08a34-3d4c-4776-9cfa-0f6ee7461674\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"0db08a34-3d4c-4776-9cfa-0f6ee7461674\">CompositionalSelector</label><div class=\"sk-toggleable__content\"><pre>CompositionalSelector()</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"11494886-9f77-4839-80fb-31a50028c1ad\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"11494886-9f77-4839-80fb-31a50028c1ad\">CLRTransform</label><div class=\"sk-toggleable__content\"><pre>CLRTransform()</pre></div></div></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>isotopes</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a5502889-a999-4093-b85d-ef04fa61b05d\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"a5502889-a999-4093-b85d-ef04fa61b05d\">CompositionalSelector</label><div class=\"sk-toggleable__content\"><pre>CompositionalSelector(inverse=True)</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"0f037166-f116-4734-9a4f-0299f0f18993\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"0f037166-f116-4734-9a4f-0299f0f18993\">LogTransform</label><div class=\"sk-toggleable__content\"><pre>LogTransform()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"26ee8206-4783-42dd-9456-1547041e94e7\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"26ee8206-4783-42dd-9456-1547041e94e7\">StandardScaler</label><div class=\"sk-toggleable__content\"><pre>StandardScaler()</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"060bea08-8eb0-4898-ba79-3ac584a2cebf\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"060bea08-8eb0-4898-ba79-3ac584a2cebf\">SVC</label><div class=\"sk-toggleable__content\"><pre>SVC(cache_size=500, class_weight='balanced', decision_function_shape='ovo',\n",
       "    probability=True)</pre></div></div></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "GridSearchCV(cv=5,\n",
       "             estimator=Pipeline(steps=[('featureunion',\n",
       "                                        FeatureUnion(transformer_list=[('abundances '\n",
       "                                                                        '--> '\n",
       "                                                                        'CLR',\n",
       "                                                                        Pipeline(steps=[('compositionalselector',\n",
       "                                                                                         CompositionalSelector()),\n",
       "                                                                                        ('clrtransform',\n",
       "                                                                                         CLRTransform())])),\n",
       "                                                                       ('isotopes',\n",
       "                                                                        Pipeline(steps=[('compositionalselector',\n",
       "                                                                                         CompositionalSelector(inverse=True)),\n",
       "                                                                                        ('logtransform',\n",
       "                                                                                         LogTransform())]))])),\n",
       "                                       ('standardscaler', StandardScaler()),\n",
       "                                       ('svc',\n",
       "                                        SVC(cache_size=500,\n",
       "                                            class_weight='balanced',\n",
       "                                            decision_function_shape='ovo',\n",
       "                                            probability=True))]),\n",
       "             n_jobs=1,\n",
       "             param_grid={'svc__C': array([  0.31622777,   1.33352143,   5.62341325,  23.71373706,\n",
       "       100.        ]),\n",
       "                         'svc__gamma': array([0.00501187, 0.01154782, 0.02660725, 0.06130558, 0.14125375])})"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_path = \"./data/basalt_classifier.pkl\"\n",
    "\n",
    "X = df[[c for c in df.columns if c!='Class']]\n",
    "y = df['Class']\n",
    "\n",
    "XX_train, XX_test, yy_train, yy_test = train_test_split(\n",
    "    X, y, stratify=y, test_size=0.3\n",
    ")\n",
    "\n",
    "try:\n",
    "    gs = joblib.load(clf_path)\n",
    "except:\n",
    "    gs = SVC_pipeline(\n",
    "        verbose=0,\n",
    "        probability=True,\n",
    "        scaler=StandardScaler(),\n",
    "        transform = FeatureUnion([\n",
    "            ('abundances --> CLR', make_pipeline(CompositionalSelector(), CLRTransform())), \n",
    "            ('isotopes', make_pipeline(CompositionalSelector(inverse=True), LogTransform()))\n",
    "                            ]),\n",
    "        param_grid=dict(\n",
    "            svc__gamma=np.logspace(-2.3, -0.85, 5),\n",
    "            svc__C=np.logspace(-0.5, 2, 5),  # limit C parameter to avoid overfitting\n",
    "        ),\n",
    "        n_jobs=1,\n",
    "        cv=5\n",
    "    )\n",
    "    gs.fit(XX_train, yy_train)\n",
    "    joblib.dump(gs, clf_path)\n",
    "gs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we've fit our classifier pipeline, we can extract the best estimator and see how it performed on the test data. While we could simply assess the overall accuracy by checking how many of the test samples it correctly classifies, we can go further and investigate how it does for each of the different tectonic settings within the test dataset. Below we've included a **confusion matrix**, which illustrates the proportion of the samples from a 'true' tectonic setting which are correctly classified (along the diagonal), and also what those which were incorrectly classified were misclassified as (off diagonals along the rows). For example, the back arc basin basalts tend to exhibit poor performance - and we can see that where they are misclassified they're commonly assigned as island arcs, intra-oceanic arcs, continental arcs and MORBs. This stems from geochemical overlap even in 20+ dimensions (as we've seen above in the extreme 2D case), which itself relates to the tectonic proximity and similarity of BAB compared to MORB and island arcs. Similarly, in some cases oceanic plateaus can be difficult to distinguish from continental flood basalts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(gs.best_estimator_, XX_test, yy_test, normalize=True);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "While we've investigated the performance of this classifier a little based on the numbers, let's quickly check what this classifier looks like relative to our Pearce plot. Below we've included a plot colored by the *predicted* tectonic setting for each sample in our dataset, and can see that by exploiting the extra dimensionality, we've managed to untangle some of the overlaps in this space. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from pyrolite.plot.color import process_color\n",
    "fig, ax = plt.subplots(1)\n",
    "pearceThNbYb(ax)\n",
    "colors = process_color(gs.best_estimator_.predict(X), cmap='tab10', alpha=0.8)['c']\n",
    "for ID in df.Class.unique():\n",
    "    subdf = df.loc[df.Class==ID, :]\n",
    "    subdf[['Nb/Yb', 'Th/Yb']].pyroplot.scatter(c=colors[df.Class==ID], ax=ax, s=8, alpha=0.8)\n",
    "    \n",
    "ax.legend([proxy_line(c=coll.get_facecolors()[0], marker='o', ls='none') for coll in ax.collections],\n",
    "          df.Class.unique(),\n",
    "          frameon=False,\n",
    "          facecolor=None)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There's an extra dimension to this model which we haven't yet explored: uncertainty. While it's relatively straightforward to envisage uncertainty for a predicted variable when it's numeric, here we have a multi-class classification problem. When we built this classifier, we did so using a method which allows for estimation of the relative probability that each sample belongs to each tectonic setting (through 'Platt scaling' of the decision surface; [Platt, 2000]; [Lin, 2007]). While this isn't the most robust way to estimate classification probability, it does provide some useful information about relative certainty.\n",
    "\n",
    "Below we've shown the estimated classification probabilities for the first few samples in our dataset across each of the eight tectonic settings, along with the true tectonic setting. Most samples are classified correctly with high certainties, for others the classification probability is distributed over multiple (typically geologically similar) settings. Where samples are incorrectly classified, it's typically with low certainty.\n",
    "\n",
    "[Platt, 2000]: \"\" \"Platt, J.C., 2000. Probabilistic Outputs for Support Vector Machines and Comparisons to Regularized Likelihood Methods, in: Advances in Large Margin Classifiers. MIT Press, pp. 61–74.\"\n",
    "\n",
    "[Lin, 2007]:https://doi.org/10/ckf84q \"Lin, H.-T., Lin, C.-J., Weng, R.C., 2007. A note on Platt’s probabilistic outputs for support vector machines. Mach Learn 68, 267–276.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col1 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col2 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col3 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col4 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col5 {\n",
       "            background-color:  #08306b;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col6 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col7 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col8 {\n",
       "            background-color:  #f2f8fd;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col1 {\n",
       "            background-color:  #f5f9fe;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col2 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col3 {\n",
       "            background-color:  #08306b;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col4 {\n",
       "            background-color:  #f6faff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col5 {\n",
       "            background-color:  #f6faff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col6 {\n",
       "            background-color:  #f6faff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col7 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col8 {\n",
       "            background-color:  #e7f0fa;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col1 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col2 {\n",
       "            background-color:  #f6faff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col3 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col4 {\n",
       "            background-color:  #f2f8fd;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col5 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col6 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col7 {\n",
       "            background-color:  #08306b;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col8 {\n",
       "            background-color:  #f6faff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col1 {\n",
       "            background-color:  #f5f9fe;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col2 {\n",
       "            background-color:  #f6faff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col3 {\n",
       "            background-color:  #08306b;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col4 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col5 {\n",
       "            background-color:  #f6faff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col6 {\n",
       "            background-color:  #f6faff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col7 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col8 {\n",
       "            background-color:  #f3f8fe;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col1 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col2 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col3 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col4 {\n",
       "            background-color:  #08306b;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col5 {\n",
       "            background-color:  #f0f6fd;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col6 {\n",
       "            background-color:  #f6faff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col7 {\n",
       "            background-color:  #eaf3fb;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col8 {\n",
       "            background-color:  #f2f7fd;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col1 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col2 {\n",
       "            background-color:  #08306b;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col3 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col4 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col5 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col6 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col7 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col8 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col1 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col2 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col3 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col4 {\n",
       "            background-color:  #08306b;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col5 {\n",
       "            background-color:  #f5fafe;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col6 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col7 {\n",
       "            background-color:  #f5fafe;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col8 {\n",
       "            background-color:  #f5f9fe;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col1 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col2 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col3 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col4 {\n",
       "            background-color:  #08306b;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col5 {\n",
       "            background-color:  #f6faff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col6 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col7 {\n",
       "            background-color:  #f7fbff;\n",
       "            color:  #000000;\n",
       "        }    #T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col8 {\n",
       "            background-color:  #f6faff;\n",
       "            color:  #000000;\n",
       "        }</style><table id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aa\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >True Tectonic Setting</th>        <th class=\"col_heading level0 col1\" >OP</th>        <th class=\"col_heading level0 col2\" >OI</th>        <th class=\"col_heading level0 col3\" >MOR</th>        <th class=\"col_heading level0 col4\" >IOA</th>        <th class=\"col_heading level0 col5\" >IA</th>        <th class=\"col_heading level0 col6\" >CFB</th>        <th class=\"col_heading level0 col7\" >CA</th>        <th class=\"col_heading level0 col8\" >BAB</th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aalevel0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col0\" class=\"data row0 col0\" >IA</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col1\" class=\"data row0 col1\" >0.070000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col2\" class=\"data row0 col2\" >0.020000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col3\" class=\"data row0 col3\" >0.050000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col4\" class=\"data row0 col4\" >0.120000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col5\" class=\"data row0 col5\" >97.220000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col6\" class=\"data row0 col6\" >0.030000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col7\" class=\"data row0 col7\" >0.120000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow0_col8\" class=\"data row0 col8\" >2.380000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aalevel0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col0\" class=\"data row1 col0\" >MOR</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col1\" class=\"data row1 col1\" >1.170000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col2\" class=\"data row1 col2\" >0.070000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col3\" class=\"data row1 col3\" >89.310000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col4\" class=\"data row1 col4\" >0.560000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col5\" class=\"data row1 col5\" >0.560000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col6\" class=\"data row1 col6\" >0.470000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col7\" class=\"data row1 col7\" >0.160000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow1_col8\" class=\"data row1 col8\" >7.700000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aalevel0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col0\" class=\"data row2 col0\" >CA</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col1\" class=\"data row2 col1\" >0.410000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col2\" class=\"data row2 col2\" >0.870000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col3\" class=\"data row2 col3\" >0.310000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col4\" class=\"data row2 col4\" >2.670000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col5\" class=\"data row2 col5\" >0.540000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col6\" class=\"data row2 col6\" >0.370000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col7\" class=\"data row2 col7\" >94.050000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow2_col8\" class=\"data row2 col8\" >0.790000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aalevel0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col0\" class=\"data row3 col0\" >MOR</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col1\" class=\"data row3 col1\" >2.490000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col2\" class=\"data row3 col2\" >1.550000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col3\" class=\"data row3 col3\" >87.150000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col4\" class=\"data row3 col4\" >1.190000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col5\" class=\"data row3 col5\" >1.750000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col6\" class=\"data row3 col6\" >1.620000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col7\" class=\"data row3 col7\" >1.300000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow3_col8\" class=\"data row3 col8\" >2.930000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aalevel0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col0\" class=\"data row4 col0\" >IOA</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col1\" class=\"data row4 col1\" >0.280000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col2\" class=\"data row4 col2\" >0.040000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col3\" class=\"data row4 col3\" >0.080000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col4\" class=\"data row4 col4\" >87.370000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col5\" class=\"data row4 col5\" >3.420000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col6\" class=\"data row4 col6\" >0.620000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col7\" class=\"data row4 col7\" >5.550000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow4_col8\" class=\"data row4 col8\" >2.640000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aalevel0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col0\" class=\"data row5 col0\" >OI</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col1\" class=\"data row5 col1\" >0.270000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col2\" class=\"data row5 col2\" >99.250000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col3\" class=\"data row5 col3\" >0.030000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col4\" class=\"data row5 col4\" >0.120000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col5\" class=\"data row5 col5\" >0.070000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col6\" class=\"data row5 col6\" >0.150000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col7\" class=\"data row5 col7\" >0.080000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow5_col8\" class=\"data row5 col8\" >0.030000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aalevel0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col0\" class=\"data row6 col0\" >IOA</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col1\" class=\"data row6 col1\" >0.040000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col2\" class=\"data row6 col2\" >0.030000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col3\" class=\"data row6 col3\" >0.070000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col4\" class=\"data row6 col4\" >96.580000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col5\" class=\"data row6 col5\" >0.830000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col6\" class=\"data row6 col6\" >0.020000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col7\" class=\"data row6 col7\" >1.080000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow6_col8\" class=\"data row6 col8\" >1.340000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aalevel0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col0\" class=\"data row7 col0\" >IOA</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col1\" class=\"data row7 col1\" >0.130000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col2\" class=\"data row7 col2\" >0.070000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col3\" class=\"data row7 col3\" >0.120000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col4\" class=\"data row7 col4\" >98.330000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col5\" class=\"data row7 col5\" >0.690000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col6\" class=\"data row7 col6\" >0.040000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col7\" class=\"data row7 col7\" >0.110000</td>\n",
       "                        <td id=\"T_f6fd8ce4_b228_11ea_afa5_00216ae618aarow7_col8\" class=\"data row7 col8\" >0.510000</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x21667fb98c8>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "samples = np.random.choice(np.arange(XX_test.shape[0]), 8, replace=False)\n",
    "probs = (\n",
    "    pd.DataFrame(\n",
    "        data=gs.best_estimator_.predict_proba(XX_test.iloc[samples]),\n",
    "        columns=gs.best_estimator_.classes_,\n",
    "    )\n",
    "    * 100\n",
    ")\n",
    "probs[\"True Tectonic Setting\"] = yy_test.iloc[samples].tolist()\n",
    "table = probs[probs.columns[::-1]].round(2)\n",
    "table.style.background_gradient(cmap=\"Blues\", axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### What can we learn?\n",
    "\n",
    "This example provides one some insight into the relative certainty with which we can classify modern basaltic rocks given a high-dimensional dataset. Overall we can approach the tectonic discrimination task with a relatively high degree of accuracy (and we can distinguish Ocean Island Basalts very well!), if we have all the data, and especially if we're using more than one sample to do so. It also highlights that in some cases the nature of source regions, magmatic and alteration processes results in some degree of 'overlap' meaning we can't always distinguish tectonic setting even for fresh rocks. \n",
    "\n",
    "By building probabilistic classifiers, we can use the multi-class classification probability outputs to better understand the model, the dataset that it is built on, and where our samples might fit in. We can provide some constraints for the maximal certainty with which we can make predictions overall, and also on a sample-by-sample basis.\n",
    "\n",
    "The next example notebook ([Data Visualization](./012_datavis.ipynb)) illustrates some methods for using these outputs to visualize the relative certainty in these multi-class classification problems, as well as visualizations which attempt to compress the dimensionality of the dataset yet preserve the overall structure.\n",
    "\n",
    "#### Limits: What We Haven't Accounted For\n",
    "\n",
    "What we don't have in this model is any accounting for the fact that the target classes may have changed with time, that there may be classes or setting which existed in the past which are no longer present (or, alternatively, that some of these classes didn't exist in the past). Further, we've made a classifier based on a relatively large set of elements and isotopes. But what if we don't have all this data, or some it is unreliable - as is the case for older deformed and altered rocks. What does this reduction in useful dimensionality do to our chances of accurately distinguishing the tectonic settings? We'll expect a decrease in overall performance as we take out potentially unreliable or confounding fluid-mobile elements and isotopes which are otherwise useful in fresher modern rocks. Notably, the effect will be different for each of the tectonic setting classes - some of these are distinguishable by a set of a few elements and isotopes (we expect this is partially the case for ocean island basalts, which have distinct isotope systematics)! These effects are testable through repeating part of the process above, but for brevity we won't continue with this here.\n",
    "\n",
    "Another aspect we haven't accounted for here is dataset shift - which in our case principally relates to i) the distribution of input variables (our elements and isotopes) which our model is trained on, and ii) 'concept shift' - or 'geological evolution' as it relates to the composition of basaltic (or otherwise) magma generated in a specific tectonic setting (in addition to the note above about tectonic settings changing!). There are methods for assessing the shift in independent variables or covariate shift (which itself is analogous to identifying where we're extrapolating our model beyond a reasonable threshold): for example, Kullback-Liebler divergence from a statistics point of view, or building a classifier to discriminate the input dataset from the training dataset from a machine learning point of view.). However, when it comes to concept shift, we know to expect it (it's part of one of our original questions..), and addressing it is challenging!\n",
    "\n",
    "<!--\n",
    "\n",
    "### Limits for Applying Models: Dataset Shift\n",
    "\n",
    "* Covariate shift (shift in independent variables) - e.g. extrapolating model to data which is significantly different to the data it has been trained with\n",
    "* Prior probability shift (e.g. imbalanced classes; distribution of predictor different between training and testing sets)\n",
    "* Concept shift (~ non-stationarity) - something fundamental about the relationship between independent variables and predictor has changed (e.g geological time & evolution).\n",
    "\n",
    "#### Identifying Covariate Shift\n",
    "\n",
    "* A statistical approach: Kullback–Leibler divergence; useful for lower dimensions\n",
    "* Which features are shifted? Multivariate vs individual features - implications of differences likely depend on models, but multivariate shift (e.g. data with a differing covariance matrix) could be a particular issue.\n",
    "# Choose a threshold at which covariate shift is sufficiently great that results have to be excluded\n",
    "-->\n",
    "\n",
    "#### Applying Models as Research Tools\n",
    "\n",
    "In a practical sense, classifiers like this can be used to extract samples from larger databases for closer inspection, and in some cases are best used as tools to refine our focus and highlight interesting samples and features!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "------\n",
    "\n",
    "### Index\n",
    "\n",
    "| [**Abstract**](./00_overview.ipynb) | **Introduction**                                                    | [**Examples**](./00_overview.ipynb#Examples)            | [**Tools**](./00_overview.ipynb#Leveraging-the-Scientific-Python-Ecosystem) | \n",
    "|:-----|:-----|:-----|:-----|\n",
    "|  | [Minerals Exploration](./00_overview.ipynb#An-Evolving-Role-of-Geochemistry-in-Mineral-Exploration)  | [Classification](./011_classification.ipynb) |  |\n",
    "|  | [Data Driven Geochem](./00_overview.ipynb#Data-Driven-Geochemistry) | [Data Visualization](./012_datavis.ipynb) |  |"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}