teamName : Koder

koder.ipynb

@@ -0,0 +1 @@
+{"metadata":{"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"name":"python","version":"3.10.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"IMPORTING NECESSARY MODULES","metadata":{}},{"cell_type":"code","source":"import pandas as pd\nimport numpy as np\nfrom sklearn.linear_model import LinearRegression\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import mean_squared_error, r2_score\nfrom sklearn.compose import ColumnTransformer\nfrom sklearn.pipeline import Pipeline\nfrom sklearn.preprocessing import OneHotEncoder\nfrom sklearn.impute import SimpleImputer\nfrom sklearn.experimental import enable_iterative_imputer\nfrom sklearn.impute import IterativeImputer\nimport seaborn as sns\nimport matplotlib.pyplot as plt","metadata":{"execution":{"iopub.status.busy":"2023-10-26T16:40:16.001800Z","iopub.execute_input":"2023-10-26T16:40:16.003728Z","iopub.status.idle":"2023-10-26T16:40:16.013156Z","shell.execute_reply.started":"2023-10-26T16:40:16.003669Z","shell.execute_reply":"2023-10-26T16:40:16.011744Z"},"trusted":true},"execution_count":7,"outputs":[]},{"cell_type":"markdown","source":"/kaggle/input/machine-learning-101/ML101_train_dataset.csvReading csv using pandas files and preprocessing the given data to remove nan values and inconsistent(negative) values.","metadata":{}},{"cell_type":"code","source":"df = pd.read_csv('/kaggle/input/machine-learning-101/ML101_train_dataset.csv')\n\nnan_counts = df.isna().sum(axis=1)\ntotal_columns = len(df.columns)-1\n\nthreshold = 1/3\n\ndf = df.drop(nan_counts[nan_counts>=total_columns*threshold].index,axis=0)\ndef putNaN(x):\n    if x < 0:\n        return np.nan\n    else:\n        return x\n\nnumeric_columns = []\nfor i in df.columns:\n    if(i==\"Gender\" or i==\"LifeStyle\"):\n        continue\n    numeric_columns.append(i)\n\n\ndf[numeric_columns] = df[numeric_columns].applymap(putNaN)\ndf.dropna(subset=['Gender'], how='all', inplace=True)\ndf.dropna(subset=['LifeStyle'], how='all', inplace=True)\n\nvalid_lifestyles = ['Good', 'Bad', 'Average', 'Great']\ndf = df[df['LifeStyle'].isin(valid_lifestyles)]\n\ndf[numeric_columns] = df[numeric_columns].interpolate(method='linear', axis=0)\n\ntrain_data=df.copy() # preprocessed training data to be used for classification","metadata":{"execution":{"iopub.status.busy":"2023-10-26T16:40:16.026528Z","iopub.execute_input":"2023-10-26T16:40:16.027066Z","iopub.status.idle":"2023-10-26T16:40:17.365961Z","shell.execute_reply.started":"2023-10-26T16:40:16.027020Z","shell.execute_reply":"2023-10-26T16:40:17.364333Z"},"trusted":true},"execution_count":8,"outputs":[]},{"cell_type":"markdown","source":"Code for prediction of blood pressure (systolic and diastolic) using regression","metadata":{}},{"cell_type":"code","source":"\ntarget_variable1,target_variable2 = 'Systolic BP' , 'Diastolic BP'\nfeatures = numeric_columns.copy()\n\n\nfeatures.remove(target_variable1)\nfeatures.remove(target_variable2)\n\ncategorical_column = ['Gender']\ndf['Gender'] = df['Gender'].map({'Male': 0, 'Female': 1})\nfeatures.append('Gender')\nX_train1,X_test1,y_train1,y_test1 = train_test_split(df[features],df[target_variable1],test_size=0.00001,random_state=42)\nX_train1 = np.array(X_train1)\nX_test1 = np.array(X_test1)\ny_train1 = np.array(y_train1)\ny_test1 = np.array(y_test1)\ndef regression_parameters(x,y):\n    Z = []\n\n    for i in range(len(x)):\n        Z.append(x[i])\n    \n    w = np.dot((np.dot(np.linalg.inv(np.dot(np.transpose(Z),Z)),np.transpose(Z))),np.transpose(y))\n\n    return w\n\ndef multiple_regression(x_train,y_train,x_test):\n    w = regression_parameters(x_train,y_train)\n    y_predict = []\n    for i in range(len(x_test)):\n        y_predict.append(np.dot(w,np.transpose(x_test[i])))\n    \n    return y_predict\n\nX_train2,X_test2,y_train2,y_test2 = train_test_split(df[features],df[target_variable2],test_size=0.00001,random_state=42)\nX_train2 = np.array(X_train2)\nX_test2 = np.array(X_test2)\ny_train2 = np.array(y_train2)\ny_test2 = np.array(y_test2)\ndef regression_parameters(x,y):\n    Z = []\n\n    for i in range(len(x)):\n        Z.append(x[i])\n    \n    w = np.dot((np.dot(np.linalg.inv(np.dot(np.transpose(Z),Z)),np.transpose(Z))),np.transpose(y))\n\n    return w\n\ndef multiple_regression(x_train,y_train,x_test):\n    w = regression_parameters(x_train,y_train)\n    y_predict = []\n    for i in range(len(x_test)):\n        y_predict.append(np.dot(w,np.transpose(x_test[i])))\n    \n    return y_predict\n\ntest_data = pd.read_csv('/kaggle/input/machine-learning-101/ML101_dataset_test_feature.csv')\nnumeric_columns.remove('Systolic BP')\nnumeric_columns.remove('Diastolic BP')\ntest_data[numeric_columns] = test_data[numeric_columns].applymap(putNaN)\ntest_data.dropna(subset=['Gender'], how='all', inplace=True)\n# test_data.dropna(subset=['LifeStyle'], how='all', inplace=True)\ntest_data['Gender'] = test_data['Gender'].map({'Male': 0, 'Female': 1})\n\nX_test = test_data[features].values\n\ny_predict_systolic = multiple_regression(X_train1, y_train1, X_test)\n\n\ny_predict_diastolic = multiple_regression(X_train2, y_train2, X_test)\n\nresults = pd.DataFrame({'Systolic BP': y_predict_systolic, 'Diastolic BP': y_predict_diastolic})\n\n\nresults.to_csv('predicted_blood_pressure.csv', index=False)","metadata":{"execution":{"iopub.status.busy":"2023-10-26T16:40:17.369268Z","iopub.execute_input":"2023-10-26T16:40:17.370942Z","iopub.status.idle":"2023-10-26T16:40:18.449369Z","shell.execute_reply.started":"2023-10-26T16:40:17.370869Z","shell.execute_reply":"2023-10-26T16:40:18.448223Z"},"trusted":true},"execution_count":9,"outputs":[]},{"cell_type":"markdown","source":"Code for Classification of given test data in terms of lifestyle","metadata":{}},{"cell_type":"code","source":"attr_name=[\"Average Daily Steps\",\"Hours of Sleep\",\"Caloric Intake\",\"Age\",\"Gender\",\"Height\",\"Weight\",\"Cholesterol level\",\"Blood Sugar level\"]\n# train_data=pd.read_csv(\"preprocessed_data.csv\")\ntrain_data=train_data.drop([\"Systolic BP\",\"Diastolic BP\"], axis=1)\n\ngr_by_gender=train_data.groupby(\"Gender\")\nmale_data=gr_by_gender.get_group(\"Male\")\nfemale_data=gr_by_gender.get_group(\"Female\")\n\ntest_data=pd.read_csv(\"/kaggle/input/machine-learning-101/ML101_dataset_test_feature.csv\")\n# test_data=x_test\ntest_data_wot_gender=test_data.drop([\"Gender\"], axis=1)\n\nattr_name=[\"Average Daily Steps\",\"Hours of Sleep\",\"Caloric Intake\",\"Age\",\"Gender\",\"Height\",\"Weight\",\"Cholesterol level\",\"Blood Sugar level\"]\nn=len(attr_name)\nclass_name=[\"Bad\",\"Average\",\"Good\",\"Great\"]\n\n# CLASSIFICATION FOR MALE\n\nmale_data=male_data.drop([\"Gender\"], axis=1)\ny_male=male_data.iloc[:,8:]\nm_data_sz=len(y_male)\n\n# separating each class data\n\ngrouped_data = male_data.groupby(\"LifeStyle\")\nm_c1=grouped_data.get_group(\"Bad\").iloc[:,:8]\nm_c2=grouped_data.get_group(\"Average\").iloc[:,:8]\nm_c3=grouped_data.get_group(\"Good\").iloc[:,:8]\nm_c4=grouped_data.get_group(\"Great\").iloc[:,:8]\n\n# parameter estimation for each class of male\nclass_c=[m_c1,m_c2,m_c3,m_c4]\nmale_mean_estimated_matrix=[]\nmale_cov_estimated_matrix=[]\n\nfor i in range(4):\n    c_i=(class_c[i])\n    mean_vec=np.array(c_i.mean(axis=0,numeric_only=True))\n    male_mean_estimated_matrix.append(mean_vec)\n    tmp=np.array(c_i).transpose()\n    cov_vec=np.cov(tmp)\n    male_cov_estimated_matrix.append(cov_vec)\n\nmale_prior_prob_c=[]\nmale_prior_prob_c.append(len(m_c1)/m_data_sz)\nmale_prior_prob_c.append(len(m_c2)/m_data_sz)\nmale_prior_prob_c.append(len(m_c3)/m_data_sz)\nmale_prior_prob_c.append(len(m_c4)/m_data_sz)\n\n# CLASSIFICATION FOR FEMALE\n\nfemale_data=female_data.drop([\"Gender\"], axis=1)\ny_female=female_data.iloc[:,8:]\nf_data_sz=len(y_female)\n\n# separating each class data\n\ngrouped_data = female_data.groupby(\"LifeStyle\")\nf_c1=grouped_data.get_group(\"Bad\").iloc[:,:8]\nf_c2=grouped_data.get_group(\"Average\").iloc[:,:8]\nf_c3=grouped_data.get_group(\"Good\").iloc[:,:8]\nf_c4=grouped_data.get_group(\"Great\").iloc[:,:8]\n\n# parameter estimation for each class of male\nclass_c=[f_c1,f_c2,f_c3,f_c4]\nfemale_mean_estimated_matrix=[]\nfemale_cov_estimated_matrix=[]\n\nfor i in range(4):\n    c_i=(class_c[i])\n    mean_vec=np.array(c_i.mean(axis=0,numeric_only=True))\n    female_mean_estimated_matrix.append(mean_vec)\n    tmp=np.array(c_i).transpose()\n    cov_vec=np.cov(tmp)\n    female_cov_estimated_matrix.append(cov_vec)\n\nfemale_prior_prob_c=[]\nfemale_prior_prob_c.append(len(f_c1)/f_data_sz)\nfemale_prior_prob_c.append(len(f_c2)/f_data_sz)\nfemale_prior_prob_c.append(len(f_c3)/f_data_sz)\nfemale_prior_prob_c.append(len(f_c4)/f_data_sz)\n\ny_test_output=[]\n\n# FUNCTION FOR MAHALNOBIS DISTANCE\ndef md(x,mu,sig):\n    x1=np.array(x-mu)\n    x2=np.linalg.inv(sig)\n    a=np.matmul(x1.transpose(),x2)\n    b=np.matmul(a,x1)\n    return b\n\n# FUNCTION FOR CLASSIFICATION OF MALE TEST DATA\ndef male_classifier(x):\n    prob_ci=[]\n    for i in range(4):\n        prob_ci.append(md(x,male_mean_estimated_matrix[i],male_cov_estimated_matrix[i]))\n    ind=np.argmin(prob_ci)\n    y_test_output.append(class_name[ind])\n\n# FUNCTION FOR CLASSIFICATION OF FEMALE TEST DATA\ndef female_classifier(x):\n    prob_ci=[]\n    for i in range(4):\n        prob_ci.append(md(x,male_mean_estimated_matrix[i],male_cov_estimated_matrix[i]))\n    ind=np.argmin(prob_ci)\n    y_test_output.append(class_name[ind])\n\nfor i in range(len(test_data.index)):\n    x_wt_g=list(np.array(test_data.iloc[i:i+1,:])[0])\n    x=list(np.array(test_data_wot_gender.iloc[i:i+1,:])[0])\n    if(x_wt_g[4]=='Male'):\n        male_classifier(x)\n    else:\n        female_classifier(x)\n\nf_results = pd.DataFrame({'LifeStyle': y_test_output})\n\n\nf_results.to_csv('predicted_lifestyle.csv', index=False)","metadata":{"execution":{"iopub.status.busy":"2023-10-26T16:40:18.451089Z","iopub.execute_input":"2023-10-26T16:40:18.451995Z","iopub.status.idle":"2023-10-26T16:40:24.632724Z","shell.execute_reply.started":"2023-10-26T16:40:18.451921Z","shell.execute_reply":"2023-10-26T16:40:24.631321Z"},"trusted":true},"execution_count":10,"outputs":[]}]}