machine_learning/03_regression_chemistry.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Python 3.10"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Linear regression with scikit-learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/scipy/__init__.py:146: UserWarning: A NumPy version >=1.17.3 and <1.25.0 is required for this version of SciPy (detected version 1.25.1\n",
      "  warnings.warn(f\"A NumPy version >={np_minversion} and <{np_maxversion}\"\n"
     ]
    }
   ],
   "source": [
    "import math\n",
    "import statistics\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.linear_model import LinearRegression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Statistical analysis of data obtained in the instrumental analysis laboratory course.\n",
    "#### The data just below are data to generate the calibration curve, and data from samples analyzed in the laboratory."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the construction of the analytical curve (regression), we will need the absorbance data in 5 different concentrations, the data we obtained are the intensities of the light when passed through the sample, and the intensity of the incident light. From these data we can easily calculate the absorbance through the equation:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$ Absorbance = -log\\frac{Io}{I} $"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Function to calculate absorbance\n",
    "absorbance = lambda I0, I: -math.log10(I0/I)\n",
    "\n",
    "# Absorbance data to generate the regression in mg/L\n",
    "x_data = {\n",
    "    \"label\": \"ABS\",\n",
    "    \"info\": \"ABS of Fe\",\n",
    "    \"data\": [\n",
    "        absorbance(1.02, 1.27), \n",
    "        absorbance(0.99, 1.53), \n",
    "        absorbance(0.97, 1.91), \n",
    "        absorbance(0.97, 2.41), \n",
    "        absorbance(0.96, 3.03)\n",
    "    ]\n",
    "}\n",
    "\n",
    "# Concentration data to generate the regression in mg/L\n",
    "y_data = {\n",
    "    \"label\": \"Concentration\",\n",
    "    \"info\": \"Concentration of the Fe curve\",\n",
    "    \"data\": [1, 2, 3, 4, 5]\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Transform our data to pass to our linear model\n",
    "X = [[i] for i in x_data[\"data\"]]\n",
    "y = y_data[\"data\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression()"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Instanciate and fit our model\n",
    "model = LinearRegression()\n",
    "model.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "tags": []
   },
   "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>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Equation</td>\n",
       "      <td>Concentration = 9.8576 x ABS + 0.0961</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>R-squared</td>\n",
       "      <td>0.999683</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           0                                      1\n",
       "0   Equation  Concentration = 9.8576 x ABS + 0.0961\n",
       "1  R-squared                               0.999683"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_summary = {\n",
    "    \"Equation\": f\"{y_data['label']} = {model.coef_[0]:.4f} x {x_data['label']} + {model.intercept_:.4f}\",\n",
    "    \"R-squared\": model.score(X, y)\n",
    "}\n",
    "\n",
    "summary_df = pd.DataFrame(model_summary.items())\n",
    "summary_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Data to plot line\n",
    "x_line = np.linspace(np.min(X), np.max(X), 1000)\n",
    "y_line = [model.predict([[i]])[0] for i in x_line]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Configure plot\n",
    "plt.style.use(\"seaborn\")\n",
    "plt.figure(figsize=(10,6))\n",
    "plt.title(f\"{x_data['label']} x {y_data['label']}\", fontsize=20)\n",
    "plt.xlabel(x_data[\"label\"], fontsize=20)\n",
    "plt.ylabel(y_data[\"label\"], fontsize=20)\n",
    "\n",
    "# Plot line\n",
    "plt.plot(x_line, y_line, color='red', zorder=1)\n",
    "\n",
    "# Plot points\n",
    "plt.scatter(x_data[\"data\"], y_data[\"data\"], color='black', zorder=2)\n",
    "\n",
    "# Show plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Making inference"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "Here we have 3 measurements of the same sample to perform the inference."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "measures = {\n",
    "    \"label\": \"ABS\",\n",
    "    \"units_of_measure\": \"mg/L\",\n",
    "    \"data\": [\n",
    "        absorbance(0.96, 1.96),\n",
    "        absorbance(0.96, 1.96),\n",
    "        absorbance(0.96, 1.97)\n",
    "    ]\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Predict\n",
    "predicts = [model.predict([[i]])[0] for i in measures[\"data\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "mean = statistics.mean(predicts)\n",
    "std = statistics.stdev(predicts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Concentration of sample: 3.1590 mg/L\n",
      "Standard deviation: 0.0126.\n"
     ]
    }
   ],
   "source": [
    "print(f\"Concentration of sample: {mean:.4f} {measures['units_of_measure']}\")\n",
    "print(f\"Standard deviation: {std:.4f}.\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "metadata": {
   "interpreter": {
    "hash": "073f00d5cf68edf6a49db4eabf702ed0e6456bf6c99dc117d2b536a857e46b3a"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}