Add files via upload

The final version for assignment 1.

assignment 1.ipynb

@@ -0,0 +1,184 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas \n",
+    "import numpy\n",
+    "from sklearn.naive_bayes import GaussianNB\n",
+    "from sklearn.model_selection import train_test_split\n",
+    "from sklearn.metrics import accuracy_score\n",
+    "from sklearn.metrics import confusion_matrix\n",
+    "from sklearn.metrics import classification_report\n",
+    "import math\n",
+    "\n",
+    "# Read .csv data from local\n",
+    "# The original data is downloaded from: https://drive.google.com/file/d/1hQjDotqV6ZrGOTnQQ8ANor5zp3L04D6M/view?usp=sharing\n",
+    "# To test in another computer, please adjust the path below.\n",
+    "\n",
+    "path = r'D:\\Uni\\SAKI\\SAKI Exercise 1 - Transaction Classification - Data Set.csv'\n",
+    "df = pandas.read_csv(path,delimiter = \";\", index_col = 0)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# The features of one type are then converted into numbers, e.g. for the type weather, [sunny, rainning, sunny] will be converted into [0, 1, 0]\n",
+    "\n",
+    "def convert (mylist):\n",
+    "    temp = mylist.copy()\n",
+    "    for i in range(len(temp)):\n",
+    "        temp[i] = i\n",
+    "    return temp"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# After analysing the data, it is to know that only in df.iloc[:,i], where i equals 0, 6, and 7, remain emplty value.\n",
+    "# In the place where i equals 0, it is a math. nan, and in the place where i equals 6 or 7, it is an empty object.\n",
+    "# Therefore, I reset all of the empty value into 0.\n",
+    "\n",
+    "for i in range(len(df.dtypes)-1):\n",
+    "    if numpy.any(df.iloc[:,i].isnull())==True:\n",
+    "\n",
+    "        for j in range(len(df.iloc[:,0])):\n",
+    "            if i == 0 and math.isnan(df.iloc[j,0]):\n",
+    "                df.iloc[j,0] = 0\n",
+    "            if i == 6 and df.iloc[j,6] is numpy.nan:\n",
+    "                df.iloc[j,6] = 0\n",
+    "            if i == 7 and df.iloc[j,7] is numpy.nan:\n",
+    "                df.iloc[j,7] = 0"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {
+    "scrolled": false
+   },
+   "outputs": [],
+   "source": [
+    "# In features_list, dinstict features of every type (ecept \"label\") will be stored into one list, which then will be appended together into the features_list array.\n",
+    "# In features_index, the distinct features of every type will be converted into numbers.\n",
+    "\n",
+    "features_list = []\n",
+    "features_index = []\n",
+    "for i in range(len(df.dtypes)-1):\n",
+    "    features_list.append(list(set(df.iloc[:,i])))\n",
+    "    features_index.append(convert(features_list[i]))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# The distinct features of \"label\" will be stored into labels_list, and the converted value will be stored into labels_index\n",
+    "\n",
+    "labels_list = list(set(df.iloc[:,len(df.dtypes)-1]))\n",
+    "labels_index = convert(labels_list)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# According to the different features in different types, and their corresponding number, the whole dataframe is \"converted\" into numbers\n",
+    "\n",
+    "features = numpy.array([[0]*(len(df.dtypes)-1) for i in range(len(df.iloc[:,0]))])\n",
+    "labels = []\n",
+    "for i in range(len(df.dtypes)):\n",
+    "    for j in range(len(df.iloc[:,0])):\n",
+    "        if i != len(df.dtypes)-1:\n",
+    "            place = features_list[i].index(df.iloc[j,i])\n",
+    "            features[j][i] = features_index[i][place]\n",
+    "        else:\n",
+    "            place = labels_list.index(df.iloc[j,i])\n",
+    "            labels.append(labels_index[place])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "classification_report: \n",
+      "                   precision    recall  f1-score   support\n",
+      "\n",
+      "         private       0.20      1.00      0.33         1\n",
+      "standardOfLiving       0.92      0.69      0.79        16\n",
+      "          income       1.00      1.00      1.00         3\n",
+      "         finance       1.00      0.89      0.94         9\n",
+      "         leisure       0.73      0.89      0.80        27\n",
+      "          living       1.00      0.29      0.44         7\n",
+      "\n",
+      "        accuracy                           0.78        63\n",
+      "       macro avg       0.81      0.79      0.72        63\n",
+      "    weighted avg       0.85      0.78      0.78        63\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "# Split the dataset into a 70%-30% train-test-process. The trained model will be used to predicte and compared with the test data.\n",
+    "# Result will be present in classification report.\n",
+    "\n",
+    "features_train, features_test, labels_train, labels_test = train_test_split(features, labels, test_size=0.3,random_state=0)\n",
+    "\n",
+    "data = {\"data\":features_train,\"target\":labels_train}\n",
+    "gnb = GaussianNB()\n",
+    "gnb = gnb.fit(data[\"data\"],data[\"target\"])\n",
+    "\n",
+    "\n",
+    "prob_pred = gnb.predict_proba(features_test)\n",
+    "pred = gnb.predict(features_test)\n",
+    "print(\"classification_report: \\n\", classification_report(labels_test, pred, target_names = labels_list))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "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": "3.8.5"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}