{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"# 1️⃣ Introduction\n\n<span style=\"color:brown;font-family:Verdana;letter-spacing:0.5px\">In Kaggle's first unsupervised clustering challengee, we have been given a dataset where each row belongs to a particular cluster. Job is to predict the cluster each row belongs to. </span>\n\n<span style=\"color:brown;font-family:Verdana;letter-spacing:0.5px\">Unsupervised learning is a type of algorithm that learns patterns from untagged data.</span>\n1. <span style=\"color:brown;font-family:Verdana;letter-spacing:0.5px\">Clustering: Clustering is a process of group similar instances together, identified as Clusters. It is extremely helpful for data analysis, customer segmentation, recommender system, search engines, image segmentation, and so many</span>\n\n![cluster-system.png](attachment:2e1b32d7-3c9a-4356-b163-82fe62da57f6.png)\n\n\nReferences:\n1. https://www.kaggle.com/code/ambrosm/tpsjul22-gaussian-mixture-cluster-analysis\n1. https://www.kaggle.com/code/ravi20076/tpsjul22-eda-baseline\n1. https://www.kaggle.com/code/kartushovdanil/tps-jul-22-advanced-2-sol","metadata":{},"attachments":{"2e1b32d7-3c9a-4356-b163-82fe62da57f6.png":{"image/png":"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"}}},{"cell_type":"code","source":"import numpy as np\nimport pandas as pd\n\nimport matplotlib.pyplot as plt\nfrom matplotlib.pyplot import figure\nimport seaborn as sns\n%matplotlib inline\n\nfrom sklearn.cluster import KMeans\nfrom sklearn.decomposition import PCA\nfrom sklearn.preprocessing import StandardScaler,RobustScaler,PowerTransformer\nfrom sklearn.metrics import silhouette_score\nfrom sklearn.mixture import BayesianGaussianMixture\n\nimport tensorflow as tf\nfrom keras.models import Model\nfrom keras.layers import Dense, Input\n\nfrom scipy.stats import shapiro\nfrom termcolor import colored\nfrom scipy.stats import chi2\n\nfrom tqdm.notebook import tqdm;\nfrom gc import collect;\nnp.random.seed(10);","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-07-27T07:05:20.727488Z","iopub.execute_input":"2022-07-27T07:05:20.728014Z","iopub.status.idle":"2022-07-27T07:05:31.590952Z","shell.execute_reply.started":"2022-07-27T07:05:20.727903Z","shell.execute_reply":"2022-07-27T07:05:31.589948Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"data_df = pd.read_csv('../input/tabular-playground-series-jul-2022/data.csv')","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:05:31.592711Z","iopub.execute_input":"2022-07-27T07:05:31.593296Z","iopub.status.idle":"2022-07-27T07:05:32.915883Z","shell.execute_reply.started":"2022-07-27T07:05:31.593264Z","shell.execute_reply":"2022-07-27T07:05:32.914586Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"data_df.head(5)","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:05:32.917443Z","iopub.execute_input":"2022-07-27T07:05:32.917925Z","iopub.status.idle":"2022-07-27T07:05:32.962517Z","shell.execute_reply.started":"2022-07-27T07:05:32.917886Z","shell.execute_reply":"2022-07-27T07:05:32.961236Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"data_df.info()","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:05:32.965416Z","iopub.execute_input":"2022-07-27T07:05:32.966308Z","iopub.status.idle":"2022-07-27T07:05:33.001320Z","shell.execute_reply.started":"2022-07-27T07:05:32.966265Z","shell.execute_reply":"2022-07-27T07:05:33.000065Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"data_df.iloc[:, 1:-1].describe().T.style.background_gradient()","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:05:33.003192Z","iopub.execute_input":"2022-07-27T07:05:33.003806Z","iopub.status.idle":"2022-07-27T07:05:33.298682Z","shell.execute_reply.started":"2022-07-27T07:05:33.003760Z","shell.execute_reply":"2022-07-27T07:05:33.297453Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"<span style=\"font-family:cursive;font-size:18px;\">💡 Observations:</span>\n* <span style=\"padding: 10px;color:blue;font-family:Verdana;letter-spacing:0.5px\">There is no null data, so we dont need to worry about it.</span>\n* <span style=\"padding: 10px;color:blue;font-family:Verdana;letter-spacing:0.5px\">There are no duplicate records.</span>\n* <span style=\"padding: 10px;color:blue;font-family:Verdana;letter-spacing:0.5px\">All the columns are in numbers either int or float. No categorical data.</span>\n* <span style=\"padding: 10px;color:blue;font-family:Verdana;letter-spacing:0.5px\">Few columns do have outliers, considering we need to cluster based on disnance, we will need to do Standardization.</span>","metadata":{"execution":{"iopub.status.busy":"2022-07-04T03:41:29.28563Z","iopub.execute_input":"2022-07-04T03:41:29.286782Z","iopub.status.idle":"2022-07-04T03:41:29.297985Z","shell.execute_reply.started":"2022-07-04T03:41:29.286724Z","shell.execute_reply":"2022-07-04T03:41:29.294139Z"}}},{"cell_type":"markdown","source":"# 2️⃣ Visualization","metadata":{}},{"cell_type":"code","source":"features = [x for x in data_df.columns if x !='id']","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:05:33.300283Z","iopub.execute_input":"2022-07-27T07:05:33.300985Z","iopub.status.idle":"2022-07-27T07:05:33.306779Z","shell.execute_reply.started":"2022-07-27T07:05:33.300936Z","shell.execute_reply":"2022-07-27T07:05:33.305603Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Generating correlation analysis without transforms:-\nfor method in tqdm(['pearson', 'spearman', 'kendall']):\n    fig, ax= plt.subplots(1,1, figsize= (20,15));\n    _corr = data_df[features].corr(method=method);\n    sns.heatmap(_corr, annot= True, fmt= '.0%', linewidth= 1.5, center= True, cmap= 'Spectral_r',\n                cbar= False, linecolor= 'white', mask = np.triu(np.ones_like(_corr)),ax= ax);\n    ax.set_title(f\"\\n{method.capitalize()} correlation plot before transforms\\n\", \n                 color= 'tab:blue', fontsize= 12);\n    plt.tight_layout();\n    plt.yticks(rotation= 0);\n    plt.xticks(rotation= 90);\n    plt.show();\n    del _corr;\n    collect();\ncollect();","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:05:33.308715Z","iopub.execute_input":"2022-07-27T07:05:33.309207Z","iopub.status.idle":"2022-07-27T07:05:52.532525Z","shell.execute_reply.started":"2022-07-27T07:05:33.309161Z","shell.execute_reply":"2022-07-27T07:05:52.531319Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"ncols = 6\nfor i, f in enumerate(features):\n    if i % ncols == 0: \n        if i > 0: plt.show()\n        plt.figure(figsize=(25, 2))\n        if i == 0: plt.suptitle('Data Symmetry check', fontsize=20, y=1.02)\n    plt.subplot(1, ncols, i % ncols + 1)\n    plt.hist(data_df[f],bins=100)\n    plt.xlabel(f)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:05:52.533964Z","iopub.execute_input":"2022-07-27T07:05:52.534326Z","iopub.status.idle":"2022-07-27T07:06:00.650327Z","shell.execute_reply.started":"2022-07-27T07:05:52.534294Z","shell.execute_reply":"2022-07-27T07:06:00.649139Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 3️⃣ Feature Engineering\n\n### 📏Shapiro-Wilk test \n(Copied from [TORCH ME](https://www.kaggle.com/code/kartushovdanil/tps-jul-22-advanced-2-sol))\n\n<span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">The Shapiro-Wilk test is a statistical test used to check if a continuous variable follows a normal distribution.</span> \n* <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">The null hypothesis (H0) states that the variable is normally distributed, </span> \n* <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">The alternative hypothesis (H1) states that the variable is NOT normally distributed.</span> \n\n<span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">The Shapiro–Wilk test tests the null hypothesis that a sample x1, ..., xn came from a normally distributed population. The test statistic is</span> \n\n![shapiro.jpg](attachment:e4c8b54f-82af-4901-ab9d-f11b416bd05c.jpg)\n\n<span style=\"padding: 8px;color:brown;font-family:Verdana;letter-spacing:0.5px\">where\nx : is the ith order statistic, i.e., the *i*th-smallest number in the sample;\nx mean: (x1+...+xn)/n is the sample mean.</span> ","metadata":{},"attachments":{"e4c8b54f-82af-4901-ab9d-f11b416bd05c.jpg":{"image/jpeg":"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"}}},{"cell_type":"code","source":"# Univariate normality test\nfor col in data_df.columns:\n    stat, p_value = shapiro(data_df[col])\n    alpha = 0.05    # significance level\n    if p_value > alpha: \n        result = colored('Accepted', 'green')  \n    else:\n        result = colored('Rejected','red')        \n    print('Feature: {}\\t p_value: {}\\t Hypothesis: {}'.format(col,round(p_value, 3), result))","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:06:00.652420Z","iopub.execute_input":"2022-07-27T07:06:00.653782Z","iopub.status.idle":"2022-07-27T07:06:00.919038Z","shell.execute_reply.started":"2022-07-27T07:06:00.653735Z","shell.execute_reply":"2022-07-27T07:06:00.917890Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"Id = data_df['id']\ndata_df = data_df.drop(['id'], axis=1)","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:06:00.922996Z","iopub.execute_input":"2022-07-27T07:06:00.923368Z","iopub.status.idle":"2022-07-27T07:06:00.934880Z","shell.execute_reply.started":"2022-07-27T07:06:00.923321Z","shell.execute_reply":"2022-07-27T07:06:00.933660Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"int_cols = [i for i in data_df.columns if data_df[i].dtype == int]\nfloat_cols = [i for i in data_df.columns if data_df[i].dtype == float]","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:06:00.936966Z","iopub.execute_input":"2022-07-27T07:06:00.937394Z","iopub.status.idle":"2022-07-27T07:06:00.950500Z","shell.execute_reply.started":"2022-07-27T07:06:00.937335Z","shell.execute_reply":"2022-07-27T07:06:00.949537Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## 🛠 Standardization:\n\n<span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">Standardization is scaling technique, values are centered around the mean with a unit standard deviation.It uses below formula,</span>\n\n> Xstnd = (X - Xmean) / Xstandard_deviation\n\n**Pros:**\n* <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">Standardization handles outliers issues without requiring prior knowledge of what the reasonable range is by linearly scaling input.</span>\n* <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">Values are not restricted to any range.</span>\n* <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">It is good for data following Gaussian distribution.</span>\n\n\n**<span style=\"font-family:Verdana;letter-spacing:0.5px\">Why Standardization is important for KMeans?</span>**\n\n<span style=\"color:brown;font-family:Verdana;letter-spacing:0.5px\">KMeans is Distance-based Algorithm, it uses the distance between data points to determine their similarity, that's why they are the most impacted algorithms if features are not in the same range.</span><br>\n<span style=\"color:brown;font-family:Verdana;letter-spacing:0.5px\">We use different distance-based techniques like Euclidean or Manhattan, Features will have a lower impact if their values are lower. Ex Salary (which in ranges in thousands) Vs Total Experience (which ranges between 0-20).</span>","metadata":{}},{"cell_type":"code","source":"scaler = PowerTransformer()\ndata_scaled = scaler.fit_transform(data_df)\ndata_scaled_df=pd.DataFrame(data_scaled, index=data_df.index, columns=data_df.columns)\ndata_scaled_df.head()","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:06:00.951961Z","iopub.execute_input":"2022-07-27T07:06:00.952416Z","iopub.status.idle":"2022-07-27T07:06:04.697866Z","shell.execute_reply.started":"2022-07-27T07:06:00.952380Z","shell.execute_reply":"2022-07-27T07:06:04.696690Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### 🛠 PCA\n\n<span style=\"color:brown;font-family:Verdana;letter-spacing:0.5px\">Principal Component Analysis (PCA) is an unsupervised linear transformation technique that is widely used across different fields, most prominently for feature extraction and dimensionality reduction. Other popular applications of PCA include exploratory data analyses and de-noising of signals in stock market trading, and the analysis of genome data and gene expression levels in the field of bioinformatics.</span>\n\n<span style=\"color:brown;font-family:Verdana;letter-spacing:0.5px\">PCA helps us to identify patterns in data based on the correlation between features. In a nutshell, PCA aims to find the directions of maximum variance in high-dimensional data and projects it onto a new subspace with equal or fewer dimensions than the original one.</span>\n\n**Algorithm Step by Step:**\n1. <span style=\"color:brown;font-family:Verdana;font-size:13px\">Standardize the d-dimensional dataset.</span>\n1. <span style=\"color:brown;font-family:Verdana;font-size:13px\">Construct the covariance matrix.</span>\n1. <span style=\"color:brown;font-family:Verdana;font-size:13px\">Decompose the covariance matrix into its eigenvectors and eigenvalues.</span>\n1. <span style=\"color:brown;font-family:Verdana;font-size:13px\">Sort the eigenvalues by decreasing order to rank the corresponding eigenvectors.</span>\n1. <span style=\"color:brown;font-family:Verdana;font-size:13px\">Select k eigenvectors which correspond to the k largest eigenvalues, where k is the dimensionality of the new feature subspace (k ≤ d).</span>\n1. <span style=\"color:brown;font-family:Verdana;font-size:13px\">Construct a projection matrix W from the “top” k eigenvectors.</span>\n1. <span style=\"color:brown;font-family:Verdana;font-size:13px\">Transform the d-dimensional input dataset X using the projection matrix W to obtain the new k-dimensional feature subspace.</span>","metadata":{}},{"cell_type":"code","source":"# Compute the PCA\npca = PCA(n_components=3)\npca.fit(data_scaled_df)\npca_df = pca.transform(data_scaled_df)","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:06:04.699176Z","iopub.execute_input":"2022-07-27T07:06:04.699524Z","iopub.status.idle":"2022-07-27T07:06:05.272939Z","shell.execute_reply.started":"2022-07-27T07:06:04.699493Z","shell.execute_reply":"2022-07-27T07:06:05.271640Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 4️⃣ Model Selection\n\n### 🛠 K Means Algorithm:\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">K-Means algorithm groups similar data points together and discover underlying patterns. To achieve this object K-Means looks for a fixed number of clusters(k) in a dataset.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">Cluster points of data points are aggregated together because of certain similarities.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">K-Means algorithm identifies k numbers of centroids, and then allocates every data point to the nearest clusters while keeping the centroid as small as possible.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">This algorithm has 'Means' as it averages data.</span>\n\n![k means.png](attachment:8d09ca14-d02a-4e19-9f03-22d7d7d8ce9f.png)\n\n**Centroids:** <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">It imaginary or real locations represent the center of the cluster.\nEvery data point is allocated to each of the clusters by reducing the in-cluster sum of squares.</span>\n\n**Inertia:**\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">It measures how well a dataset was clustered by k-means.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">It is calculated by measuring the distance between each point and its centroid, squaring this distance, and summing these squares acrocss one cluster.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">A good model is one with low inertia and lower number of clusters(k), but as number of clusters increases inertia decreases.</span>\n\n**Elbow Method:**\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">This method is used to find the optimal number of cluster (k). It finds a point where inertia begins to decrease slowly.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">It generally makes shape of elbow while drawing relation between inertia and number of clusters, thats why known as Elbow Method.</span>\n\n![elbow.png](attachment:1b9fb4b8-e6a3-462f-bec3-a8a291d20c4f.png)\n\n**Algorithm Step by Step:**\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">It randomly places k centroids for initial clusters.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">Assign each data point to its nearest centroid.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">Update centroid location based on the location of data points.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">Repeat step 2 & 3 untill centroids stabilize.</span>\n\n**Advantages:**\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">Its relatively simple to implement.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">It scales to large data sets.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">It easily adapts new example.</span>\n\n**Limitations:**\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">Need to specify the number of clusters, it's a hassle to specify that number.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">Need to run the algorithm several times to avoid sub-optimal solutions.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">It doesn't perform well if clusters have varying sizes, different densities, or non-spherical shapes.</span>\n1. <span style=\"padding: 10px;color:brown;font-family:Verdana;letter-spacing:0.5px\">Centroids can be dragged by outliers or even may create their own cluster. </span>","metadata":{},"attachments":{"8d09ca14-d02a-4e19-9f03-22d7d7d8ce9f.png":{"image/png":"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"},"1b9fb4b8-e6a3-462f-bec3-a8a291d20c4f.png":{"image/png":"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"}}},{"cell_type":"code","source":"# fitting multiple k-means algorithms and storing the values in an empty list\nSSE = []\nfor cluster in range(1,12):\n    kmeans = KMeans(n_clusters = cluster, init='k-means++')\n    kmeans.fit(pca_df)\n    SSE.append(kmeans.inertia_)    \n\n# converting the results into a dataframe and plotting them\nframe = pd.DataFrame({'Cluster':range(1,12), 'SSE':SSE})\nplt.figure(figsize=(12,6))\nplt.plot(frame['Cluster'], frame['SSE'], marker='o')\nplt.xlabel('Number of clusters')\nplt.ylabel('Inertia')","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:06:05.279681Z","iopub.execute_input":"2022-07-27T07:06:05.283249Z","iopub.status.idle":"2022-07-27T07:07:04.629850Z","shell.execute_reply.started":"2022-07-27T07:06:05.283178Z","shell.execute_reply":"2022-07-27T07:07:04.628540Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"kmeans = KMeans(algorithm='auto', copy_x=True, init='k-means++', max_iter=300, n_clusters=7, n_init=25,verbose=0)\nkmeans = kmeans.fit(pca_df)\ny_pred_x = kmeans.labels_","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:07:04.631881Z","iopub.execute_input":"2022-07-27T07:07:04.632384Z","iopub.status.idle":"2022-07-27T07:07:22.247504Z","shell.execute_reply.started":"2022-07-27T07:07:04.632310Z","shell.execute_reply":"2022-07-27T07:07:22.246402Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### 🛠 Bayesian Gaussian Mixture:\n<span style=\"color:brown;font-family:Verdana;letter-spacing:0.5px\">It is probabilistics unsupervised model that assumes that the instances were generated from a mixture of several gaussian distributions whose paramters are unknown.\nGood thing with Bayesian Gaussian Mixture is, **its capable of giving weights equal to zero to unnecessary clusters.**\nIt is variational Bayesian estimation of a Gaussian mixture.\nWe just need to setnumber of clusters greater than expected optimal numbers of clusters.</span>\n\n<span style=\"color:brown;font-family:Verdana;letter-spacing:0.5px\">Ex. If you believe this problem will have 6 clusters, you can give any number higher than 7, it will assign 0 value to all clusters above 6.</span>","metadata":{}},{"cell_type":"code","source":"best_data =['f_07','f_08', 'f_09', 'f_10','f_11', 'f_12', 'f_13', 'f_22','f_23', 'f_24', 'f_25','f_26','f_27', 'f_28']","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:07:22.248787Z","iopub.execute_input":"2022-07-27T07:07:22.249094Z","iopub.status.idle":"2022-07-27T07:07:22.253908Z","shell.execute_reply.started":"2022-07-27T07:07:22.249058Z","shell.execute_reply":"2022-07-27T07:07:22.253167Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"bgm = BayesianGaussianMixture(n_components=7,covariance_type='full',random_state=42,max_iter=100, n_init=10)\ny_pred_x = bgm.fit_predict(data_scaled_df[best_data])","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:07:22.255031Z","iopub.execute_input":"2022-07-27T07:07:22.255512Z","iopub.status.idle":"2022-07-27T07:23:03.582994Z","shell.execute_reply.started":"2022-07-27T07:07:22.255481Z","shell.execute_reply":"2022-07-27T07:23:03.581638Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"np.round(bgm.weights_,2)","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:23:03.584420Z","iopub.execute_input":"2022-07-27T07:23:03.585192Z","iopub.status.idle":"2022-07-27T07:23:03.592001Z","shell.execute_reply.started":"2022-07-27T07:23:03.585148Z","shell.execute_reply":"2022-07-27T07:23:03.591214Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#Create a  DataFrame with the passengers ids and our prediction\nsubmission = pd.DataFrame({'Id':Id,'Predicted':y_pred_x})","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:23:03.593030Z","iopub.execute_input":"2022-07-27T07:23:03.593589Z","iopub.status.idle":"2022-07-27T07:23:03.605312Z","shell.execute_reply.started":"2022-07-27T07:23:03.593550Z","shell.execute_reply":"2022-07-27T07:23:03.604262Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sns.histplot(submission['Predicted'], stat=\"percent\", discrete=True)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:23:03.606796Z","iopub.execute_input":"2022-07-27T07:23:03.607253Z","iopub.status.idle":"2022-07-27T07:23:03.922738Z","shell.execute_reply.started":"2022-07-27T07:23:03.607221Z","shell.execute_reply":"2022-07-27T07:23:03.921434Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"data_scaled_df[\"class\"] = y_pred_x","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:23:03.924130Z","iopub.execute_input":"2022-07-27T07:23:03.925050Z","iopub.status.idle":"2022-07-27T07:23:03.930375Z","shell.execute_reply.started":"2022-07-27T07:23:03.925012Z","shell.execute_reply":"2022-07-27T07:23:03.929376Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"f,ax = plt.subplots(figsize=(25,50))\nfor n,feature in enumerate(int_cols):\n    plt.subplot(8,3,n+1)\n    sns.kdeplot(data=data_scaled_df, x=feature, hue=\"class\", palette=sns.color_palette(\"hls\", 7));","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:23:03.931850Z","iopub.execute_input":"2022-07-27T07:23:03.932170Z","iopub.status.idle":"2022-07-27T07:23:12.612089Z","shell.execute_reply.started":"2022-07-27T07:23:03.932141Z","shell.execute_reply":"2022-07-27T07:23:12.610805Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"f,ax = plt.subplots(figsize=(25,50))\nfor n,feature in enumerate(float_cols):\n    plt.subplot(8,3,n+1)\n    sns.kdeplot(data=data_scaled_df, x=feature, hue=\"class\", palette=sns.color_palette(\"hls\", 7));","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:23:12.614051Z","iopub.execute_input":"2022-07-27T07:23:12.614503Z","iopub.status.idle":"2022-07-27T07:23:40.058118Z","shell.execute_reply.started":"2022-07-27T07:23:12.614460Z","shell.execute_reply":"2022-07-27T07:23:40.056933Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"y = submission.Predicted","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:23:40.059727Z","iopub.execute_input":"2022-07-27T07:23:40.060067Z","iopub.status.idle":"2022-07-27T07:23:40.065523Z","shell.execute_reply.started":"2022-07-27T07:23:40.060037Z","shell.execute_reply":"2022-07-27T07:23:40.064219Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"prop_cycle = plt.rcParams['axes.prop_cycle']\n\n# PCA projection, random drawing order of points\nc = [prop_cycle.by_key()['color'][i % 10] for i in y]\n\nplt.figure(figsize=(8, 8))\nplt.scatter(pca_df[:,0], pca_df[:,1], s=1, label=f\"Cluster {i}\", c=c)\nplt.xlabel('PCA[0]')\nplt.ylabel('PCA[1]')\nplt.legend()\nplt.title('PCA projection')\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:23:40.067051Z","iopub.execute_input":"2022-07-27T07:23:40.067419Z","iopub.status.idle":"2022-07-27T07:23:43.157350Z","shell.execute_reply.started":"2022-07-27T07:23:40.067382Z","shell.execute_reply":"2022-07-27T07:23:43.156048Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission.head(5)","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:23:43.158883Z","iopub.execute_input":"2022-07-27T07:23:43.159223Z","iopub.status.idle":"2022-07-27T07:23:43.168371Z","shell.execute_reply.started":"2022-07-27T07:23:43.159193Z","shell.execute_reply":"2022-07-27T07:23:43.167445Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission.to_csv('submission.csv',index=False)","metadata":{"execution":{"iopub.status.busy":"2022-07-27T07:23:43.169705Z","iopub.execute_input":"2022-07-27T07:23:43.170255Z","iopub.status.idle":"2022-07-27T07:23:43.337697Z","shell.execute_reply.started":"2022-07-27T07:23:43.170220Z","shell.execute_reply":"2022-07-27T07:23:43.336545Z"},"trusted":true},"execution_count":null,"outputs":[]}]}