Covid 19 Visualisation and Prediction .ipynb

{"cells":[{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load in \n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n\n# Input data files are available in the \"../input/\" directory.\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n\n# Any results you write to the current directory are saved as output.","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"import numpy as np \nimport matplotlib.pyplot as plt \nimport matplotlib.colors as mcolors\nimport pandas as pd \nimport random\nimport math\nimport time\nimport os\nfrom sklearn.linear_model import LinearRegression, BayesianRidge\nfrom sklearn.model_selection import RandomizedSearchCV, train_test_split\nfrom sklearn.preprocessing import PolynomialFeatures\nfrom sklearn.svm import SVR\nfrom sklearn.metrics import mean_squared_error, mean_absolute_error\nimport datetime\nimport operator \nplt.style.use('fivethirtyeight')\n%matplotlib inline ","execution_count":2,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"confirmed_df = pd.read_csv('/kaggle/input/novel-corona-virus-2019-dataset/time_series_covid_19_confirmed.csv')\ndeaths_df = pd.read_csv('/kaggle/input/novel-corona-virus-2019-dataset/time_series_covid_19_deaths.csv')\nrecoveries_df = pd.read_csv('/kaggle/input/novel-corona-virus-2019-dataset/time_series_covid_19_recovered.csv')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"confirmed_df.dropna(axis=1)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"columns = confirmed_df.keys()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Extract the case datewise so that we may computed Mortality rate etc."},{"metadata":{"trusted":true},"cell_type":"code","source":"confirmed = confirmed_df.loc[:,columns[4]:columns[-1]]\nrecoveries = recoveries_df.loc[:, columns[4]: columns[-1]]\ndeaths = deaths_df.loc[:, columns[4]: columns[-1]]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"dates = confirmed.keys()\nworld_cases = []\ntotal_deaths = [] \nmortality_rate = []\ntotal_recovered = [] \nchina_cases=[]\nspain_cases=[]\nusa_cases=[]\nindia_cases=[]\nitaly_cases=[]\nfor i in dates:\n    confirmed_sum = confirmed[i].sum()\n    death_sum = deaths[i].sum()\n    recovered_sum = recoveries[i].sum()\n    world_cases.append(confirmed_sum)\n    total_deaths.append(death_sum)\n    mortality_rate.append(death_sum/confirmed_sum)\n    total_recovered.append(recovered_sum)\n    \n    china_cases.append(confirmed_df[confirmed_df['Country/Region']=='China'][i].sum())\n    spain_cases.append(confirmed_df[confirmed_df['Country/Region']=='Spain'][i].sum())\n    usa_cases.append(confirmed_df[confirmed_df['Country/Region']=='US'][i].sum())\n    india_cases.append(confirmed_df[confirmed_df['Country/Region']=='India'][i].sum())\n    italy_cases.append(confirmed_df[confirmed_df['Country/Region']=='Italy'][i].sum())\n    \n\"\"\"    \nprint(india_cases)\n\"\"\"","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Daywise increase"},{"metadata":{"trusted":true},"cell_type":"code","source":"def daywise_increase(data):\n    d=[]\n    for i in range(len(data)):\n        if i==0:\n            d.append(data[0])\n        else:\n            d.append(data[i]-data[i-1])\n    return d\n        \nworld_daily_increase= daywise_increase(world_cases)\nIndia_daily_increase= daywise_increase(india_cases)\nItaly_daily_increase= daywise_increase(italy_cases)\nUS_daily_increase= daywise_increase(usa_cases)\nChina_daily_increase= daywise_increase(china_cases)\nSpain_daily_increase = daywise_increase(spain_cases)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"days_since_1_22 = np.array([i for i in range (len(dates))]).reshape(-1,1)\nworld_cases = np.array(world_cases).reshape(-1,1)\ntotal_deaths = np.array(total_deaths).reshape(-1,1)\ntotal_recovered = np.array(total_recovered).reshape(-1,1)\n","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Convert to datetime"},{"metadata":{"trusted":true},"cell_type":"code","source":"days_in_future = 10\nfuture_forecast = np.array([i for i in range(len(dates)+days_in_future)]).reshape(-1,1)\nadjusted_dates = future_forecast[:-10]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"start = '1/22/2020'\nstart_date = datetime.datetime.strptime(start, '%m/%d/%Y')\nfuture_forcast_dates = []\nfor i in range(len(future_forecast)):\n    future_forcast_dates.append((start_date + datetime.timedelta(days=i)).strftime('%m/%d/%Y'))\n    \n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"X_train_confirmed, X_test_confirmed, y_train_confirmed, y_test_confirmed = train_test_split(days_since_1_22, world_cases, test_size=0.10, shuffle=False) ","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Using Linear Regression, SVM and Ridge Regression for prediction"},{"metadata":{},"cell_type":"markdown","source":"# ***SVM***"},{"metadata":{"trusted":true},"cell_type":"code","source":"\"\"\"\nkernel = ['poly', 'sigmoid', 'rbf']\nc = [0.0001,0.001, 0.01, 0.1, 1, 10]\ngamma = [0.01, 0.1, 1]\nepsilon = [0.01, 0.1, 1]\nshrinking = [True, False]\n\nsvm_grid = {'kernel': kernel, 'C': c, 'gamma' : gamma, 'epsilon': epsilon, 'shrinking' : shrinking}\nsvm = SVR()\nsvm_search = RandomizedSearchCV(svm, svm_grid, scoring='neg_mean_squared_error', cv=3, return_train_score=True, n_jobs=-1, n_iter=40, verbose=1)\nsvm_search.fit(X_train_confirmed, y_train_confirmed)\n\"\"\"\nsvm_confirmed = SVR(shrinking=True, kernel='poly',gamma=0.01, epsilon=1,degree=8, C=0.1)\nsvm_confirmed.fit(X_train_confirmed, y_train_confirmed)\nsvm_pred = svm_confirmed.predict(future_forecast)\n\n\"\"\"\"\nkernel = ['poly', 'sigmoid', 'rbf']\nc = [0.01, 0.1, 1, 10]\ngamma = [0.01, 0.1, 1]\nepsilon = [0.01, 0.1, 1]\nshrinking = [True, False]\n\n\nsvm = SVR()\nsvm_search = RandomizedSearchCV(svm, svm_grid, scoring='neg_mean_squared_error', cv=3, return_train_score=True, n_jobs=-1, n_iter=40, verbose=1)\nsvm_search.fit(X_train_confirmed, y_train_confirmed)\n\"\"\"","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"svm_search.best_params_","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"svm_predict","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"svm_test_pred = svm_confirmed.predict(X_test_confirmed)\nplt.plot(y_test_confirmed)\nplt.plot(svm_test_pred)\nplt.legend(['Test Data', 'SVM Predictions'])\nprint('MAE:', mean_absolute_error(svm_test_pred, y_test_confirmed))\nprint('MSE:',mean_squared_error(svm_test_pred, y_test_confirmed))","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Linear Regression\n"},{"metadata":{"trusted":true},"cell_type":"code","source":"linear_model= LinearRegression(normalize= True, fit_intercept = True)\nlinear_model.fit(X_train_confirmed, y_train_confirmed)\n\ntest_linear = linear_model.predict(X_test_confirmed);\nlinear_pred= linear_model.predict(future_forecast);\nprint('MAE:', mean_absolute_error(test_linear, y_test_confirmed))\nprint('MSE:', mean_squared_error(test_linear, y_test_confirmed))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(linear_model.coef_)\nprint(linear_model.intercept_)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Polynomial Regression**"},{"metadata":{"trusted":true},"cell_type":"code","source":"poly = PolynomialFeatures(degree=6)\npoly_X_train_confirmed = poly.fit_transform(X_train_confirmed)\npoly_X_test_confirmed = poly.fit_transform(X_test_confirmed)\npoly_future_forecast = poly.fit_transform(future_forecast)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"linear_model= LinearRegression(normalize= True, fit_intercept = False)\nlinear_model.fit(poly_X_train_confirmed, y_train_confirmed)\n\ntest_linear = linear_model.predict(poly_X_test_confirmed);\nlinear_pred= linear_model.predict(poly_future_forecast);\nprint('MAE:', mean_absolute_error(test_linear, y_test_confirmed))\nprint('MSE:', mean_squared_error(test_linear, y_test_confirmed))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"plt.plot(y_test_confirmed)\nplt.plot(test_linear)\nplt.legend(['Test Data', 'Polynomial Regression Predictions'])","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Bayesian Grid"},{"metadata":{"trusted":true},"cell_type":"code","source":"\n\ntol = [1e-4, 1e-3, 1e-2]\nalpha_1 = [1e-7, 1e-6, 1e-5, 1e-4]\nalpha_2 = [1e-7, 1e-6, 1e-5, 1e-4]\nlambda_1 = [1e-7, 1e-6, 1e-5, 1e-4]\nlambda_2 = [1e-7, 1e-6, 1e-5, 1e-4]\n\nbayesian_grid = {'tol': tol, 'alpha_1': alpha_1, 'alpha_2' : alpha_2, 'lambda_1': lambda_1, 'lambda_2' : lambda_2}\n\nbayesian = BayesianRidge(fit_intercept=False, normalize=True)\nbayesian_search = RandomizedSearchCV(bayesian, bayesian_grid, scoring='neg_mean_squared_error', cv=3, return_train_score=True, n_jobs=-1, n_iter=40, verbose=1)\nbayesian_search.fit(poly_X_train_confirmed, y_train_confirmed)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"bayesian_search.best_params_","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"bayesian_search.best_estimator_","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"bayesian_predict","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"bayesian_confirmed = bayesian_search.best_estimator_\ntest_bayesian_pred = bayesian_confirmed.predict(poly_X_test_confirmed)\nbayesian_pred = bayesian_confirmed.predict(poly_future_forcast)\nprint('MAE:', mean_absolute_error(test_bayesian_pred, y_test_confirmed))\nprint('MSE:',mean_squared_error(test_bayesian_pred, y_test_confirmed))\nplt.plot(y_test_confirmed)\nplt.plot(test_bayesian_pred)\nplt.legend(['Test Data', 'Bayesian Ridge Polynomial Predictions'])\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"\nplt.plot(figsize=(16,9))\nplt.plot(adjusted_dates, world_cases)\nplt.title('# of Coronavirus Cases Over Time', size=30)\nplt.xlabel('Days Since 1/22/2020', size=30, color='red')\nplt.ylabel('No of cases', size=30, color= 'red')\nplt.xticks(size=15)\nplt.yticks(size=15)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"adjusted_dates = adjusted_dates.reshape(1,-1)[0]\nplt.figure(figsize=(16, 9))\nplt.bar(adjusted_dates, world_daily_increase)\nplt.title('World Daily Increases in Confirmed Cases', size=30)\nplt.xlabel('Days Since 1/22/2020', size=30)\nplt.ylabel('# of Cases', size=30)\nplt.xticks(size=20)\nplt.yticks(size=20)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"\nplt.figure(figsize=(16, 9))\nplt.bar(adjusted_dates, china_cases)\nplt.title('China Daily Increase', size=30)\nplt.xlabel('Days Since 1/22/2020', size=30)\nplt.ylabel('# of Cases', size=30)\nplt.xticks(size=20)\nplt.yticks(size=20)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"adjusted_dates = adjusted_dates.reshape(1,-1)[0]\nplt.figure(figsize=(16, 9))\nplt.bar(adjusted_dates, spain_cases)\nplt.title('Spain Daily Increase', size=30)\nplt.xlabel('Days Since 1/22/2020', size=30)\nplt.ylabel('# of Cases', size=30)\nplt.xticks(size=20)\nplt.yticks(size=20)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"adjusted_dates = adjusted_dates.reshape(1,-1)[0]\nplt.figure(figsize=(16, 9))\nplt.bar(adjusted_dates, india_cases)\nplt.title('India Daily Increase', size=30)\nplt.xlabel('Days Since 1/22/2020', size=30)\nplt.ylabel('# of Cases', size=30)\nplt.xticks(size=20)\nplt.yticks(size=20)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"adjusted_dates = adjusted_dates.reshape(1,-1)[0]\nplt.figure(figsize=(16, 9))\nplt.bar(adjusted_dates, usa_cases)\nplt.title('US Daily Increase', size=30)\nplt.xlabel('Days Since 1/22/2020', size=30)\nplt.ylabel('# of Cases', size=30)\nplt.xticks(size=20)\nplt.yticks(size=20)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"adjusted_dates = adjusted_dates.reshape(1,-1)[0]\nplt.figure(figsize=(16, 9))\nplt.bar(adjusted_dates, italy_cases)\nplt.title('Italy Daily Increase', size=30)\nplt.xlabel('Days Since 1/22/2020', size=30)\nplt.ylabel('# of Cases', size=30)\nplt.xticks(size=20)\nplt.yticks(size=20)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"plt.figure(figsize=(16, 9))\nplt.plot(adjusted_dates, china_cases)\nplt.plot(adjusted_dates, italy_cases)\nplt.plot(adjusted_dates, usa_cases)\nplt.plot(adjusted_dates, spain_cases)\nplt.plot(adjusted_dates, india_cases)\nplt.title('# of Coronavirus Cases', size=30)\nplt.xlabel('Days Since 1/22/2020', size=30)\nplt.ylabel('# of Cases', size=30)\nplt.legend(['China', 'Italy', 'US', 'Spain','India'], prop={'size': 20})\nplt.xticks(size=20)\nplt.yticks(size=20)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**SVM Prediction**"},{"metadata":{"trusted":true},"cell_type":"code","source":"\n\n\nplt.figure(figsize=(16, 9))\nplt.plot(adjusted_dates, world_cases)\nplt.plot(future_forecast, svm_pred, linestyle='dashed', color='purple')\nplt.title('# of Coronavirus Cases Over Time', size=30)\nplt.xlabel('Days Since 1/22/2020', size=30)\nplt.ylabel('# of Cases', size=30)\nplt.legend(['Confirmed Cases', 'SVM predictions'], prop={'size': 20})\nplt.xticks(size=20)\nplt.yticks(size=20)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"> **Poly  regression Prediction**"},{"metadata":{"trusted":true},"cell_type":"code","source":"plt.figure(figsize=(16, 9))\nplt.plot(adjusted_dates, world_cases)\nplt.plot(future_forecast, linear_pred, linestyle='dashed', color='orange')\nplt.title('# of Coronavirus Cases Over Time', size=30)\nplt.xlabel('Days Since 1/22/2020', size=30)\nplt.ylabel('# of Cases', size=30)\nplt.legend(['Confirmed Cases', 'Polynomial Regression Predictions'], prop={'size': 20})\nplt.xticks(size=20)\nplt.yticks(size=20)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Bayesian Ridge Prediction**"},{"metadata":{"trusted":true},"cell_type":"code","source":"plt.figure(figsize=(16, 9))\nplt.plot(adjusted_dates, world_cases)\nplt.plot(future_forecast, bayesian_pred, linestyle='dashed', color='green')\nplt.title('# of Coronavirus Cases Over Time', size=30)\nplt.xlabel('Time', size=30)\nplt.ylabel('# of Cases', size=30)\nplt.legend(['Confirmed Cases', 'Polynomial Bayesian Ridge Regression Predictions'], prop={'size': 20})\nplt.xticks(size=20)\nplt.yticks(size=20)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Future Predictions**"},{"metadata":{},"cell_type":"markdown","source":"**a)SVM Fut Prediction**"},{"metadata":{"trusted":true},"cell_type":"code","source":"\nprint('SVM future predictions:')\nset(zip(future_forcast_dates[-10:], np.round(svm_pred[-10:])))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"\nlinear_pred = linear_pred.reshape(1,-1)[0]\nprint('Polynomial regression future predictions:')\nset(zip(future_forcast_dates[-10:], np.round(linear_pred[-10:])))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"print('Ridge regression future predictions:')\nset(zip(future_forcast_dates[-10:], np.round(bayesian_pred[-10:])))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"mean_mortality_rate = np.mean(mortality_rate)\nplt.figure(figsize=(16, 9))\nplt.plot(adjusted_dates, mortality_rate, color='orange')\nplt.axhline(y = mean_mortality_rate,linestyle='--', color='black')\nplt.title('Mortality Rate of Coronavirus Over Time', size=30)\nplt.legend(['mortality rate', 'y='+str(mean_mortality_rate)], prop={'size': 20})\nplt.xlabel('Days Since 1/22/2020', size=30)\nplt.ylabel('Mortality Rate', size=30)\nplt.xticks(size=20)\nplt.yticks(size=20)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]}],"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":4,"nbformat_minor":4}