Skip to content

Instantly share code, notes, and snippets.

Show Gist options
  • Save abegehr/0d74bcfbc3726b6cb39270a49ee5ed3b to your computer and use it in GitHub Desktop.
Save abegehr/0d74bcfbc3726b6cb39270a49ee5ed3b to your computer and use it in GitHub Desktop.
Display the source blob
Display the rendered blob
Raw
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# COVID19 week4 SIR-Model (LeoCorona)\n",
"https://www.kaggle.com/c/covid19-global-forecasting-week-4"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"id": "Eo9ulW5s6iz5"
},
"outputs": [],
"source": [
"from sklearn.metrics import mean_squared_log_error\n",
"import warnings\n",
"warnings.filterwarnings('ignore')\n",
"import pandas as pd\n",
"from sklearn import preprocessing\n",
"import numpy as np\n",
"from scipy import integrate, optimize\n",
"import math\n",
"\n",
"predictions_total = []\n",
"actual_total = []\n",
"val_loss_dict = {}\n",
"\n",
"val_info_dict = {}\n",
"predictions_dict = {}\n",
"actuals_dict = {}\n",
"colors_dict = {}\n",
"loss_dict = {}\n",
"train_start = 0\n",
"train_end = 0\n",
"val_start = 0\n",
"val_end = 0\n",
"test_start = 0\n",
"test_end = 0\n",
"modes = [\"Confirmed Cases\", \"Fatalities\"]\n",
"method = \"SIR\"\n",
"dynamic_start_day = False"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"id": "bL4IrSy_-xqw"
},
"outputs": [],
"source": [
"test = pd.read_csv(\"../input/covid19-global-forecasting-week-4/test.csv\", parse_dates=[\"Date\"])\n",
"train = pd.read_csv(\"../input/covid19-global-forecasting-week-4/train.csv\")\n",
"submission = pd.read_csv(\"../input/covid19-global-forecasting-week-4/submission.csv\")\n",
"all_data = train.copy()\n",
"# Create date columns\n",
"all_data['Date'] = pd.to_datetime(all_data['Date'])\n",
"le = preprocessing.LabelEncoder()\n",
"all_data['Day_num'] = le.fit_transform(all_data.Date)\n",
"all_data['Day'] = all_data['Date'].dt.day\n",
"all_data['Month'] = all_data['Date'].dt.month\n",
"all_data['Year'] = all_data['Date'].dt.year"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"id": "MEX_1h2CYUyZ",
"outputId": "7bef57b8-18b5-42a3-8518-069f9849dc4b"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Cleaned country details dataset\n",
"Joined dataset\n",
"Encoded dataset\n"
]
}
],
"source": [
"# Load countries data file\n",
"world_population = pd.read_csv(\"../input/population-by-country-2020/population_by_country_2020.csv\")\n",
"\n",
"# Select desired columns and rename some of them\n",
"world_population = world_population[['Country (or dependency)', 'Population (2020)', 'Density (P/Km²)', 'Land Area (Km²)', 'Med. Age', 'Urban Pop %']]\n",
"world_population.columns = ['Country (or dependency)', 'Population (2020)', 'Density', 'Land Area', 'Med Age', 'Urban Pop']\n",
"\n",
"# Replace United States by US\n",
"world_population.loc[world_population['Country (or dependency)']=='United States', 'Country (or dependency)'] = 'US'\n",
"\n",
"# Remove the % character from Urban Pop values\n",
"world_population['Urban Pop'] = world_population['Urban Pop'].str.rstrip('%')\n",
"\n",
"# Replace Urban Pop and Med Age \"N.A\" by their respective modes, then transform to int\n",
"world_population.loc[world_population['Urban Pop']=='N.A.', 'Urban Pop'] = int(world_population.loc[world_population['Urban Pop']!='N.A.', 'Urban Pop'].mode()[0])\n",
"world_population['Urban Pop'] = world_population['Urban Pop'].astype('int16')\n",
"world_population.loc[world_population['Med Age']=='N.A.', 'Med Age'] = int(world_population.loc[world_population['Med Age']!='N.A.', 'Med Age'].mode()[0])\n",
"world_population['Med Age'] = world_population['Med Age'].astype('int16')\n",
"\n",
"print(\"Cleaned country details dataset\")\n",
"\n",
"\n",
"# Now join the dataset to our previous DataFrame and clean missings (not match in left join)- label encode cities\n",
"print(\"Joined dataset\")\n",
"all_data = all_data.merge(world_population, left_on='Country_Region', right_on='Country (or dependency)', how='left')\n",
"all_data[['Population (2020)', 'Density', 'Land Area', 'Med Age', 'Urban Pop']] = all_data[['Population (2020)', 'Density', 'Land Area', 'Med Age', 'Urban Pop']].fillna(0)\n",
"\n",
"\n",
"print(\"Encoded dataset\")\n",
"# Label encode countries and provinces. Save dictionary for exploration purposes\n",
"all_data.drop('Country (or dependency)', inplace=True, axis=1)\n",
"all_data['Country_Region'] = le.fit_transform(all_data['Country_Region'])\n",
"\n",
"number_c = all_data['Country_Region']\n",
"countries = le.inverse_transform(all_data['Country_Region'])\n",
"country_dict = dict(zip(countries, number_c)) \n",
"all_data['Province_State'].fillna(\"None\", inplace=True)\n",
"all_data['Province_State'] = le.fit_transform(all_data['Province_State'])\n",
"number_p = all_data['Province_State']\n",
"province = le.inverse_transform(all_data['Province_State'])\n",
"province_dict = dict(zip(province, number_p)) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## SIR Model"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"id": "bGq-gMq5oqUV"
},
"outputs": [],
"source": [
"class SIR:\n",
" def __init__(self, beta=0, gamma=0, fix_gamma=False):\n",
" self.beta = beta\n",
" self.gamma = gamma\n",
" self.infected_t0 = 0\n",
" self.fitted_on = np.array([])\n",
" self.fix_gamma = fix_gamma\n",
" self.fitted = False\n",
" \n",
" def ode(self, y, x, beta, gamma):\n",
" '''Defines the ODE that governs the SIRs behaviour'''\n",
" dSdt = -beta * y[0] * y[1]\n",
" dRdt = gamma * y[1]\n",
" dIdt = -(dSdt + dRdt)\n",
" return dSdt, dIdt, dRdt\n",
" \n",
" def solve_ode(self, x, beta, gamma):\n",
" '''Solves the resulting ODE to get predictions for each time step'''\n",
" return np.cumsum(integrate.odeint(self.ode, (1-self.infected_t0, self.infected_t0, 0.0), x, args=(beta, gamma))[:,1])\n",
" \n",
" def solve_ode_fixed(self, x, beta):\n",
" '''Solves the resulting ODE to get predictions for each time step'''\n",
" return np.cumsum(integrate.odeint(self.ode, (1-self.infected_t0, self.infected_t0, 0.0), x, args=(beta, self.gamma))[:,1])\n",
" \n",
" def describe(self):\n",
" assert self.fitted, \"You need to fit the model before describing it!\"\n",
" print(\"Beta: \", self.beta)\n",
" print(\"Gamma: \", self.gamma)\n",
" print(\"At t=0: \", self.infected_t0)\n",
" \n",
" plt.plot(range(1,len(self.fitted_on)+1), self.fitted_on, \"x\", label='Actual')\n",
" plt.plot(range(1,len(self.fitted_on)+1), self.predict(len(self.fitted_on)), label='Prediction')\n",
" plt.title(\"Fit of SIR model to actual\")\n",
" plt.ylabel(\"% of Population\")\n",
" plt.xlabel(\"Days\")\n",
" plt.legend()\n",
" plt.show()\n",
" \n",
" def evaluate(self, y_test):\n",
" assert self.fitted, \"You need to fit the model before evaluating it!\"\n",
" print(\"Beta: \", self.beta)\n",
" print(\"Gamma: \", self.gamma)\n",
" print(\"At t=0: \", self.infected_t0)\n",
" \n",
" y_train = self.fitted_on\n",
" l_train = len(self.fitted_on)\n",
" l_test = len(y_test)\n",
" l_all = l_train + l_test\n",
" \n",
" plt.plot(range(1, l_train + 1), y_train, \"x\", label='Actual Train')\n",
" plt.plot(range(1 + l_train, l_all + 1), y_test, \"x\", label='Actual Test')\n",
" plt.plot(range(1, l_all + 1), self.predict(l_all), label='Prediction')\n",
" plt.title(\"Fit of SIR model to actual\")\n",
" plt.ylabel(\"% of Population\")\n",
" plt.xlabel(\"Days\")\n",
" plt.legend()\n",
" plt.show()\n",
" \n",
" def fit(self, y):\n",
" '''Fits the parameters to the data, assuming the first data point is the start of the outbreak'''\n",
" if len(y) == 1: y = np.array([0, y[0]]) # SIR needs at least 2 datapoints to fit\n",
" self.infected_t0 = y[0]\n",
" x = np.array(range(1,len(y)+1), dtype=float)\n",
" self.fitted_on = y\n",
" if(self.fix_gamma):\n",
" popt, _ = optimize.curve_fit(self.solve_ode_fixed, x, y)\n",
" self.beta = popt[0]\n",
" else:\n",
" popt, _ = optimize.curve_fit(self.solve_ode, x, y, maxfev=1000)\n",
" self.beta = popt[0]\n",
" self.gamma = popt[1]\n",
" self.fitted = True\n",
" \n",
" def predict(self ,length):\n",
" '''Returns the predicted cumulated cases at each time step, assuming outbreak starts at t=0'''\n",
" #assert self.fitted, \"You need to fit the model before predicting!\"\n",
" return self.solve_ode(range(1, length+1), self.beta, self.gamma)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Data Prep"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"id": "_r7cJkVjnlsn"
},
"outputs": [],
"source": [
"unknown_countries = []\n",
"hardcoded_countries = {\n",
" \"Korea, South\": 51269000,\n",
" \"Diamond Princess\": 3711,\n",
" \"Taiwan*\": 23800000,\n",
" \"Saint Vincent and the Grenadines\": 109897,\n",
" \"Congo (Brazzaville)\":5261000,\n",
" \"Congo (Kinshasa)\":81340000,\n",
" \"Cote d'Ivoire\":24300000,\n",
" \"Czechia\": 10650000,\n",
" \"Saint Kitts and Nevis\": 55345,\n",
" \"Burma\": 53370000,\n",
" \"Kosovo\": 1831000,\n",
" \"MS Zaandam\": 1432, # cruise ship\n",
" \"West Bank and Gaza\": 4685,\n",
" \"Sao Tome and Principe\": 204327,\n",
"}\n",
"hardcoded_province = {\n",
" \"Saint Pierre and Miquelon\": 5888,\n",
" \"Bonaire, Sint Eustatius and Saba\": 25157,\n",
" \"Falkland Islands (Malvinas)\": 2840,\n",
"}\n",
"state_populations= pd.read_csv(\"../input/covid19-forecasting-metadata/region_metadata.csv\")\n",
"\n",
"def get_population(country_name, province_name=None):\n",
" if province_name:\n",
" pop = state_populations[state_populations['Province_State']==province_name]['population']\n",
" if len(pop)==0:\n",
" if province_name in hardcoded_province:\n",
" return hardcoded_province[province_name]\n",
" else:\n",
" print(f\"Warning: We have no province population data at the moment. Instead of data for {province_name}, using data for {country_name}\")\n",
" else:\n",
" return pop.iloc[0]\n",
" \n",
" if country_name in hardcoded_countries:\n",
" return hardcoded_countries[country_name]\n",
" \n",
" pop = all_data[all_data[\"Country_Region\"] == country_dict[country_name]].iloc[0][\"Population (2020)\"]\n",
" if not pop:\n",
" print(f\"population of {country_name} unknown\")\n",
" pop = 100\n",
" unknown_countries.append(country_name)\n",
" \n",
" return pop"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"id": "QELZtpCdQp8P",
"outputId": "09e65ef5-ade6-4800-da95-105f01a41fd3"
},
"outputs": [
{
"data": {
"text/plain": [
"1438116346.0"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"country_name = 'China'\n",
"all_data[all_data[\"Country_Region\"] == country_dict[country_name]].iloc[0][\"Population (2020)\"]"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"id": "tsjtS2dcCh6j",
"outputId": "fdde0367-1edd-4244-937e-c09f24723c5e"
},
"outputs": [
{
"data": {
"text/plain": [
"1438116346.0"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"country_name = 'Hubei'\n",
"all_data[all_data[\"Province_State\"] == province_dict[country_name]].iloc[0][\"Population (2020)\"]"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"id": "B3d-fcFy5bK2"
},
"outputs": [],
"source": [
"def get_country_data(country_name, province_name=None, train_split_factor=1.0):\n",
" if province_name:\n",
" confirmed_total_date_country = train[(train['Country_Region']==country_name) & (train['Province_State']==province_name)].groupby(['Date']).agg({'ConfirmedCases':['sum']})\n",
" fatalities_total_date_country = train[(train['Country_Region']==country_name) & (train['Province_State']==province_name)].groupby(['Date']).agg({'Fatalities':['sum']})\n",
" total_date_country = confirmed_total_date_country.join(fatalities_total_date_country)\n",
"\n",
" cases = total_date_country.ConfirmedCases['sum'].values\n",
" cases_normalized = total_date_country.ConfirmedCases['sum'].values / get_population(country_name, province_name)\n",
" fatalities_normalized = total_date_country.Fatalities['sum'].values / get_population(country_name, province_name)\n",
"\n",
" cases_final = cases_normalized[np.argmax(cases>0):]\n",
" fatalities_final = fatalities_normalized[np.argmax(fatalities_normalized>0):]\n",
"\n",
" cases_length = len(cases_final)\n",
" fat_length = len(fatalities_final)\n",
" cases_split = math.floor(cases_length * train_split_factor)\n",
" fat_split = math.floor(fat_length * train_split_factor)\n",
" else:\n",
" confirmed_total_date_country = train[train['Country_Region']==country_name].groupby(['Date']).agg({'ConfirmedCases':['sum']})\n",
" fatalities_total_date_country = train[train['Country_Region']==country_name].groupby(['Date']).agg({'Fatalities':['sum']})\n",
" total_date_country = confirmed_total_date_country.join(fatalities_total_date_country)\n",
"\n",
" cases = total_date_country.ConfirmedCases['sum'].values\n",
" cases_normalized = cases / get_population(country_name, province_name)\n",
" fatalities_normalized = total_date_country.Fatalities['sum'].values / get_population(country_name, province_name)\n",
"\n",
" cases_final = cases_normalized[np.argmax(cases>0):]\n",
" fatalities_final = fatalities_normalized[np.argmax(fatalities_normalized>0):]\n",
"\n",
" cases_length = len(cases_final)\n",
" fat_length = len(fatalities_final)\n",
" cases_split = math.floor(cases_length * train_split_factor)\n",
" fat_split = math.floor(fat_length * train_split_factor)\n",
" \n",
" return cases_final, fatalities_final, cases_split, fat_split, cases_length, fat_length"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Visualization Helpers"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"id": "DbvCXYEbsV_l"
},
"outputs": [],
"source": [
"import matplotlib.pyplot as plt\n",
"def visualize(val_loss_dict, val_info_dict, start=0, end=150):\n",
" fig = plt.figure(figsize=(10,2))\n",
" ax = fig.add_axes([0,0,1,1])\n",
"\n",
" loss_sorted = sorted(val_loss_dict.items(), key=lambda x: x[1], reverse=True)\n",
" print(loss_sorted[10:20])\n",
" losses = [x[1] for x in loss_sorted[start:end]]\n",
" countries = [x[0] for x in loss_sorted[start:end]]\n",
" colors = [val_info_dict[x][\"Color\"] for x in countries]\n",
" ax.bar(countries, losses, color=colors)"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"id": "gzpYg0IBtp3h"
},
"outputs": [],
"source": [
"def visualize_country(country_name, val_info_dict=val_info_dict):\n",
" info = val_info_dict[country_name]\n",
" cases_actual = info[\"Cases Actual\"]\n",
" cases_predicted = info[\"Cases Predicted\"]\n",
" cases_split = info[\"Case Split\"]\n",
" fat_actual = info[\"Fatalities Actual\"]\n",
" fat_predicted = info[\"Fatalities Predicted\"]\n",
" fat_split = info[\"Fatality Split\"]\n",
" \n",
" fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(30,15))\n",
"\n",
" ax1.plot(cases_actual, 'o')\n",
" ax1.plot(cases_predicted)\n",
" ax1.axvline(x=cases_split, color='gray', linestyle='--')\n",
" ax1.set_title(\"Fit of SIR model to global infected cases\")\n",
" \n",
" ax2.plot(fat_actual, 'o')\n",
" ax2.plot(fat_predicted)\n",
" ax2.axvline(x=fat_split, color='gray', linestyle='--')\n",
" ax2.set_title(\"Fit of SIR model to global fatalities\")\n",
" \n",
" plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Training"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"id": "6IvVbeCU7YUz"
},
"outputs": [],
"source": [
"def train_val_country(country_name, train_split_factor=1.0):\n",
" cases, fatalities, case_split, fat_split, case_length, fat_length = get_country_data(country_name, train_split_factor=train_split_factor)\n",
" cases_train = cases[0:case_split]\n",
" cases_test = cases[case_split:]\n",
" fat_train = fatalities[0:fat_split]\n",
" fat_test = fatalities[fat_split:]\n",
" \n",
" case_model = SIR()\n",
" case_model.fit(cases_train)\n",
" fat_model = SIR()\n",
" fat_model.fit(fat_train)\n",
" \n",
" cases_pred_all = case_model.predict(len(cases_train) + len(cases_test))\n",
" cases_pred_train = cases_pred_all[:case_split]\n",
" cases_pred_test = cases_pred_all[case_split:]\n",
" fat_pred_all = fat_model.predict(len(fat_train) + len(fat_test))\n",
" fat_pred_train = fat_pred_all[:fat_split]\n",
" fat_pred_test = fat_pred_all[fat_split:]\n",
" \n",
" if(sum(cases_test) > sum(cases_pred_test)):\n",
" color = \"red\"\n",
" else:\n",
" color = \"blue\"\n",
" \n",
" cases_train_val_loss = np.sqrt(mean_squared_log_error(cases_train, cases_pred_train)) if (len(cases_train) > 0) else 0\n",
" fat_train_val_loss = np.sqrt(mean_squared_log_error(fat_train, fat_pred_train)) if (len(fat_train) > 0) else 0\n",
" cases_test_val_loss = np.sqrt(mean_squared_log_error(cases_test, cases_pred_test)) if (len(cases_test) > 0) else 0\n",
" fat_test_val_loss = np.sqrt(mean_squared_log_error(fat_test, fat_pred_test)) if (len(fat_test) > 0) else 0\n",
" #print(f\"Val Loss for {country_name}: {val_loss}\")\n",
" #print(f\"Sum actual: {sum(cases_test)} Sum predicted: {sum(cases_pred_val)}\")\n",
" val_loss_dict[country_name] = cases_test_val_loss\n",
" results_dict = {\n",
" \"Country\": country_name,\n",
" \"Province\": float('nan'),\n",
" \"Case Model\": case_model,\n",
" \"Fatality Model\": fat_model,\n",
" \"Color\": color,\n",
" \"Cases Predicted\": cases_pred_all,\n",
" \"Cases Actual\": cases,\n",
" \"Fatalities Predicted\": fat_pred_all,\n",
" \"Fatalities Actual\": fatalities,\n",
" \"Cases Loss Train\": cases_train_val_loss,\n",
" \"Fatality Loss Train\": fat_train_val_loss,\n",
" \"Cases Loss Test\": cases_test_val_loss,\n",
" \"Fatality Loss Test\": fat_test_val_loss,\n",
" \"Case Split\": case_split,\n",
" \"Fatality Split\": fat_split,\n",
" \"Case length\": case_length,\n",
" \"Fatality length\": fat_length\n",
" }\n",
" return results_dict\n",
"\n",
"def train_val_province(country_name, province_name, train_split_factor=1.0):\n",
" cases, fatalities, case_split, fat_split, case_length, fat_length = get_country_data(country_name, province_name, train_split_factor=train_split_factor)\n",
" cases_train = cases[0:case_split]\n",
" cases_test = cases[case_split:]\n",
" fat_train = fatalities[0:fat_split]\n",
" fat_test = fatalities[fat_split:]\n",
" \n",
" case_model = SIR()\n",
" case_model.fit(cases_train)\n",
" fat_model = SIR()\n",
" fat_model.fit(fat_train)\n",
" \n",
" cases_pred_all = case_model.predict(len(cases_train) + len(cases_test))\n",
" cases_pred_train = cases_pred_all[:case_split]\n",
" cases_pred_test = cases_pred_all[case_split:]\n",
" fat_pred_all = fat_model.predict(len(fat_train) + len(fat_test))\n",
" fat_pred_train = fat_pred_all[:fat_split]\n",
" fat_pred_test = fat_pred_all[fat_split:]\n",
" \n",
" if(sum(cases_test) > sum(cases_pred_test)):\n",
" color = \"red\"\n",
" else:\n",
" color = \"blue\"\n",
"\n",
" cases_train_val_loss = np.sqrt(mean_squared_log_error(cases_train, cases_pred_train)) if (len(cases_train) > 0) else 0\n",
" fat_train_val_loss = np.sqrt(mean_squared_log_error(fat_train, fat_pred_train)) if (len(fat_train) > 0) else 0\n",
" cases_test_val_loss = np.sqrt(mean_squared_log_error(cases_test, cases_pred_test)) if (len(cases_test) > 0) else 0\n",
" fat_test_val_loss = np.sqrt(mean_squared_log_error(fat_test, fat_pred_test)) if (len(fat_test) > 0) else 0\n",
" #print(f\"Val Loss for {country_name}: {val_loss}\")\n",
" #print(f\"Sum actual: {sum(cases_test)} Sum predicted: {sum(cases_pred_val)}\")\n",
" val_loss_dict[province_name] = cases_test_val_loss\n",
" results_dict = {\n",
" \"Country\": country_name,\n",
" \"Province\": province_name,\n",
" \"Case Model\": case_model,\n",
" \"Fatality Model\": fat_model,\n",
" \"Color\": color,\n",
" \"Cases Predicted\": cases_pred_all,\n",
" \"Cases Actual\": cases,\n",
" \"Fatalities Predicted\": fat_pred_all,\n",
" \"Fatalities Actual\": fatalities,\n",
" \"Cases Loss Train\": cases_train_val_loss,\n",
" \"Fatality Loss Train\": fat_train_val_loss,\n",
" \"Cases Loss Test\": cases_test_val_loss,\n",
" \"Fatality Loss Test\": fat_test_val_loss,\n",
" \"Case Split\": case_split,\n",
" \"Fatality Split\": fat_split,\n",
" \"Case length\": case_length,\n",
" \"Fatality length\": fat_length\n",
" }\n",
" return results_dict"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"id": "LZynM7cQ8fUN"
},
"outputs": [],
"source": [
"country_and_provinces = {}\n",
"only_provinces = {}\n",
"only_country = []\n",
"for country in test['Country_Region'].unique():\n",
" provinces = test[test['Country_Region']==country]['Province_State'].unique()\n",
" \n",
" if len(provinces)>1:\n",
" contains_nan = False\n",
" for province in provinces:\n",
" if type(province) == float:\n",
" contains_nan = True\n",
" if contains_nan:\n",
" country_and_provinces[country] = provinces\n",
" else:\n",
" only_provinces[country] = provinces\n",
" else:\n",
" only_country.append(country)\n"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {
"id": "eoEqisvajzsY",
"outputId": "0ed82a6b-94a5-4928-8eb5-3948dddc611e"
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 184/184 [00:24<00:00, 7.53it/s]\n"
]
}
],
"source": [
"from tqdm import tqdm\n",
"\n",
"train_split_factor = 0.9\n",
"\n",
"for country in tqdm(train['Country_Region'].unique()):\n",
" #If we only need to predict for the provinces, not for the whole country\n",
" if country in only_provinces:\n",
" for province in only_provinces[country]:\n",
" val_info_dict[province] = train_val_province(country, province, train_split_factor=train_split_factor)\n",
" \n",
" #If we need to predict for the provinces and for the whole country\n",
" elif country in country_and_provinces:\n",
" for province in country_and_provinces[country]:\n",
" #For the 'nan' province value: Make predictions for the whole country\n",
" if type(province) == float:\n",
" val_info_dict[country] = train_val_country(country, train_split_factor=train_split_factor)\n",
" else:\n",
" val_info_dict[province] = train_val_province(country, province, train_split_factor=train_split_factor)\n",
" \n",
" #If we don't have any provinces for this country\n",
" elif country in only_country:\n",
" val_info_dict[country] = train_val_country(country, train_split_factor=train_split_factor)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Results"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"{'Afghanistan': 1.2849805882297129e-06,\n",
" 'Albania': 7.179673514664038e-06,\n",
" 'Algeria': 2.311680547714642e-05,\n",
" 'Andorra': 0.00011380427356961391,\n",
" 'Angola': 1.83245844824193e-08,\n",
" 'Antigua and Barbuda': 7.796286718083674e-05,\n",
" 'Argentina': 4.606806475610179e-05,\n",
" 'Armenia': 0.00027505426536992997,\n",
" 'Australian Capital Territory': 8.853711342700853e-05,\n",
" 'New South Wales': 4.997783712991756e-05,\n",
" 'Northern Territory': 2.7453079591036178e-05,\n",
" 'Queensland': 5.73788133074826e-05,\n",
" 'South Australia': 1.8425943996740522e-05,\n",
" 'Tasmania': 3.569775087839743e-05,\n",
" 'Victoria': 0.00014693735369897645,\n",
" 'Western Australia': 9.654178851669384e-05,\n",
" 'Austria': 0.0010463090882809523,\n",
" 'Azerbaijan': 1.0059204702777996e-05,\n",
" 'Bahamas': 2.1788486201736117e-06,\n",
" 'Bahrain': 0.00026743168585066656,\n",
" 'Bangladesh': 3.1835889689564354e-06,\n",
" 'Barbados': 2.2303183829751415e-05,\n",
" 'Belarus': 0.00011702311186299479,\n",
" 'Belgium': 0.0004383376194654168,\n",
" 'Belize': 2.3887580549634123e-05,\n",
" 'Benin': 4.2369033498519815e-07,\n",
" 'Bhutan': 2.680172100714255e-06,\n",
" 'Bolivia': 7.010622961859406e-07,\n",
" 'Bosnia and Herzegovina': 0.0002075848205904463,\n",
" 'Botswana': 1.732374238185675e-06,\n",
" 'Brazil': 8.453395192549809e-05,\n",
" 'Brunei': 1.9534378280180854e-05,\n",
" 'Bulgaria': 4.686678489912542e-05,\n",
" 'Burkina Faso': 1.7163778218086352e-05,\n",
" 'Burma': 7.081479220440122e-07,\n",
" 'Burundi': 1.3641049933980924e-07,\n",
" 'Cabo Verde': 5.414606538709324e-06,\n",
" 'Cambodia': 1.5944615629192908e-06,\n",
" 'Cameroon': 2.109961616750171e-05,\n",
" 'Alberta': 0.00032055529594577834,\n",
" 'British Columbia': 0.00021677780292031907,\n",
" 'Manitoba': 3.309409172838959e-05,\n",
" 'New Brunswick': 1.3194813525915369e-05,\n",
" 'Newfoundland and Labrador': 1.2413172370969668e-05,\n",
" 'Northwest Territories': 1.8142081563949042e-05,\n",
" 'Nova Scotia': 7.222807860571439e-05,\n",
" 'Ontario': 8.772414666261898e-05,\n",
" 'Prince Edward Island': 2.129581461488591e-05,\n",
" 'Quebec': 0.0014828736626624401,\n",
" 'Saskatchewan': 6.3715023688719484e-06,\n",
" 'Yukon': 4.1650957622535295e-05,\n",
" 'Central African Republic': 8.953086537405141e-07,\n",
" 'Chad': 7.407117368522066e-07,\n",
" 'Chile': 0.0004153297743491753,\n",
" 'Anhui': 2.1566689139812655e-05,\n",
" 'Beijing': 2.6229565948391174e-06,\n",
" 'Chongqing': 5.7631453832574635e-08,\n",
" 'Fujian': 7.6019198198622304e-06,\n",
" 'Gansu': 1.6451591438256645e-06,\n",
" 'Guangdong': 1.3051481502116888e-06,\n",
" 'Guangxi': 3.0594524991051327e-06,\n",
" 'Guizhou': 2.5156739997183374e-06,\n",
" 'Hainan': 3.819901166330451e-06,\n",
" 'Hebei': 4.142053536523168e-06,\n",
" 'Heilongjiang': 7.238223756184245e-06,\n",
" 'Henan': 1.2236664988679813e-05,\n",
" 'Hong Kong': 9.763761425172463e-05,\n",
" 'Hubei': 1.0548900721327478e-05,\n",
" 'Hunan': 1.3637809109980497e-05,\n",
" 'Inner Mongolia': 1.2959307213464633e-06,\n",
" 'Jiangsu': 9.609667934990638e-06,\n",
" 'Jiangxi': 2.2755236066561313e-05,\n",
" 'Jilin': 1.2317509196038182e-06,\n",
" 'Liaoning': 9.553044151037838e-07,\n",
" 'Macau': 3.534290212370054e-05,\n",
" 'Ningxia': 5.308178505254384e-09,\n",
" 'Qinghai': 6.613373443466151e-08,\n",
" 'Shaanxi': 2.5261381292458854e-08,\n",
" 'Shandong': 7.885347476129242e-06,\n",
" 'Shanghai': 3.851036574706187e-06,\n",
" 'Shanxi': 1.0783754923652749e-06,\n",
" 'Sichuan': 3.567460498739672e-07,\n",
" 'Tianjin': 2.3862196846591877e-06,\n",
" 'Tibet': 1.5620604608296038e-12,\n",
" 'Xinjiang': 7.747761139738117e-07,\n",
" 'Yunnan': 2.846683448461055e-06,\n",
" 'Zhejiang': 2.0599924368058365e-06,\n",
" 'Colombia': 3.9084086457108786e-05,\n",
" 'Congo (Brazzaville)': 8.176248984776057e-07,\n",
" 'Congo (Kinshasa)': 1.1453189268570956e-06,\n",
" 'Costa Rica': 8.688118111562011e-05,\n",
" \"Cote d'Ivoire\": 7.204477644459691e-06,\n",
" 'Croatia': 0.0002590150833213281,\n",
" 'Cuba': 4.0840849786138845e-06,\n",
" 'Cyprus': 6.243690161459026e-06,\n",
" 'Czechia': 0.0005871861150005884,\n",
" 'Faroe Islands': 0.0001305659256466765,\n",
" 'Greenland': 3.603043499316395e-05,\n",
" 'Denmark': 0.0012591971527977609,\n",
" 'Diamond Princess': 0.0028284886543598346,\n",
" 'Djibouti': 7.444796408205375e-05,\n",
" 'Dominica': 2.0942769012272434e-05,\n",
" 'Dominican Republic': 0.000162158723695759,\n",
" 'Ecuador': 7.739318760134494e-05,\n",
" 'Egypt': 1.6672291572369493e-06,\n",
" 'El Salvador': 4.699286997379525e-06,\n",
" 'Equatorial Guinea': 7.714263286201959e-06,\n",
" 'Eritrea': 1.474604984397985e-06,\n",
" 'Estonia': 0.0006738493427564092,\n",
" 'Eswatini': 3.5618437456067586e-06,\n",
" 'Ethiopia': 5.100515059015048e-08,\n",
" 'Fiji': 2.4327863474647453e-06,\n",
" 'Finland': 0.0003279808210551552,\n",
" 'French Guiana': 0.00013398949436807886,\n",
" 'French Polynesia': 8.123726486541588e-05,\n",
" 'Guadeloupe': 1.469638697756939e-05,\n",
" 'Martinique': 7.703353094394305e-05,\n",
" 'Mayotte': 2.531573584602383e-05,\n",
" 'New Caledonia': 1.1830528960676557e-06,\n",
" 'Reunion': 5.792214707737096e-06,\n",
" 'Saint Barthelemy': 0.00023406160201836092,\n",
" 'Saint Pierre and Miquelon': 1.1677838844580471e-11,\n",
" 'St Martin': 0.0002135746509242852,\n",
" 'France': 0.0010104327526954083,\n",
" 'Gabon': 7.510302496475318e-06,\n",
" 'Gambia': 2.1036881987238083e-06,\n",
" 'Georgia': 0.0010140726625794663,\n",
" 'Germany': 8.409149255787794e-05,\n",
" 'Ghana': 1.6074047176706742e-06,\n",
" 'Greece': 0.00020463048801767328,\n",
" 'Grenada': 4.85706957133777e-06,\n",
" 'Guatemala': 9.596880499085776e-07,\n",
" 'Guinea': 2.697362538223174e-06,\n",
" 'Guinea-Bissau': 9.428352208242373e-07,\n",
" 'Guyana': 1.5133697855269813e-05,\n",
" 'Haiti': 8.268039774387613e-07,\n",
" 'Holy See': 0.004411124109562615,\n",
" 'Honduras': 1.5566439546521945e-05,\n",
" 'Hungary': 2.4688742074428397e-05,\n",
" 'Iceland': 0.004561714116364317,\n",
" 'India': 2.6424478286460263e-06,\n",
" 'Indonesia': 7.544530203816871e-06,\n",
" 'Iran': 0.0007038178820593384,\n",
" 'Iraq': 2.0548941622067137e-05,\n",
" 'Ireland': 0.0016420296695800357,\n",
" 'Israel': 3.740961728137474e-05,\n",
" 'Italy': 0.0019895647310509694,\n",
" 'Jamaica': 1.8741651045458255e-06,\n",
" 'Japan': 1.441285653047277e-05,\n",
" 'Jordan': 3.144443463861551e-05,\n",
" 'Kazakhstan': 1.4400188126095315e-05,\n",
" 'Kenya': 1.9731091003428376e-06,\n",
" 'Korea, South': 0.00036406925092194334,\n",
" 'Kosovo': 9.450696026603212e-05,\n",
" 'Kuwait': 0.00021356200372572363,\n",
" 'Kyrgyzstan': 1.9122272968050875e-06,\n",
" 'Laos': 5.821354311149184e-07,\n",
" 'Latvia': 0.00021026446444712913,\n",
" 'Lebanon': 6.374786835181886e-05,\n",
" 'Liberia': 6.046388765133942e-06,\n",
" 'Libya': 5.114971608578015e-07,\n",
" 'Liechtenstein': 0.00023814207843863712,\n",
" 'Lithuania': 2.9198335717654533e-05,\n",
" 'Luxembourg': 0.00034153927536285613,\n",
" 'MS Zaandam': 0.0003046000357504367,\n",
" 'Madagascar': 1.4801136211412838e-07,\n",
" 'Malawi': 3.439503197558827e-07,\n",
" 'Malaysia': 1.547418445013923e-05,\n",
" 'Maldives': 6.935977480746684e-06,\n",
" 'Mali': 4.229233591169801e-07,\n",
" 'Malta': 6.0665659023603595e-05,\n",
" 'Mauritania': 4.297897435663005e-07,\n",
" 'Mauritius': 6.197912849760222e-05,\n",
" 'Mexico': 1.0069840701698503e-05,\n",
" 'Moldova': 0.00021774708507984954,\n",
" 'Monaco': 0.00014323627567936414,\n",
" 'Mongolia': 2.256785782348178e-06,\n",
" 'Montenegro': 9.975932465295745e-06,\n",
" 'Morocco': 2.4503481414768627e-05,\n",
" 'Mozambique': 1.8472220968626146e-07,\n",
" 'Namibia': 1.403074733748481e-06,\n",
" 'Nepal': 3.17916386856687e-07,\n",
" 'Aruba': 6.460452522259138e-05,\n",
" 'Bonaire, Sint Eustatius and Saba': 3.5715945145786376e-05,\n",
" 'Curacao': 8.319910846078059e-06,\n",
" 'Sint Maarten': 0.00020481917519548254,\n",
" 'Netherlands': 0.00032544515198131566,\n",
" 'New Zealand': 0.000212729796348943,\n",
" 'Nicaragua': 5.214825941961771e-07,\n",
" 'Niger': 1.4896733742015378e-05,\n",
" 'Nigeria': 2.1804780524167706e-07,\n",
" 'North Macedonia': 0.0002547146718233912,\n",
" 'Norway': 0.0007488766880086114,\n",
" 'Oman': 6.04687007386788e-05,\n",
" 'Pakistan': 1.3605953260894951e-05,\n",
" 'Panama': 0.0007110676844440865,\n",
" 'Papua New Guinea': 8.0581307462877e-08,\n",
" 'Paraguay': 4.611596197112679e-06,\n",
" 'Peru': 5.921097639621376e-05,\n",
" 'Philippines': 4.51311391563911e-06,\n",
" 'Poland': 0.00019926593035117531,\n",
" 'Portugal': 0.0020090052825980516,\n",
" 'Qatar': 2.814324358593336e-05,\n",
" 'Romania': 0.00023183626336175473,\n",
" 'Russia': 5.35960394170891e-05,\n",
" 'Rwanda': 5.605379737950622e-06,\n",
" 'Saint Kitts and Nevis': 3.8142097296000895e-05,\n",
" 'Saint Lucia': 1.0111810024833086e-05,\n",
" 'Saint Vincent and the Grenadines': 7.502198593195164e-05,\n",
" 'San Marino': 0.0023135838973522157,\n",
" 'Sao Tome and Principe': 4.704141603703062e-13,\n",
" 'Saudi Arabia': 7.774870833088573e-05,\n",
" 'Senegal': 7.554969102456319e-06,\n",
" 'Serbia': 0.0002712179053672993,\n",
" 'Seychelles': 1.7494044915745855e-05,\n",
" 'Sierra Leone': 1.2114273028744893e-07,\n",
" 'Singapore': 0.00033238046861956027,\n",
" 'Slovakia': 9.638337659743148e-05,\n",
" 'Slovenia': 0.0003269185428750306,\n",
" 'Somalia': 2.6961864616424584e-06,\n",
" 'South Africa': 4.556875544337852e-05,\n",
" 'South Sudan': 1.3682715018445416e-08,\n",
" 'Spain': 6.747567701671771e-05,\n",
" 'Sri Lanka': 3.0040664491785722e-06,\n",
" 'Sudan': 3.643966547857473e-07,\n",
" 'Suriname': 3.447734893930676e-07,\n",
" 'Sweden': 9.488454021709986e-05,\n",
" 'Switzerland': 0.002024124499842536,\n",
" 'Syria': 1.5834271637320346e-07,\n",
" 'Taiwan*': 1.2218999978852254e-06,\n",
" 'Tanzania': 2.3987083857060715e-07,\n",
" 'Thailand': 2.969099381962298e-06,\n",
" 'Timor-Leste': 2.5355151317515834e-06,\n",
" 'Togo': 6.121565495982997e-07,\n",
" 'Trinidad and Tobago': 3.3820625104494725e-06,\n",
" 'Tunisia': 4.963227922583118e-05,\n",
" 'Turkey': 0.0011840397491182512,\n",
" 'Alabama': 0.00011006440910276353,\n",
" 'Alaska': 0.00020733281629836502,\n",
" 'Arizona': 3.262323188611717e-05,\n",
" 'Arkansas': 0.00038769623833965393,\n",
" 'California': 0.0001028189104813389,\n",
" 'Colorado': 0.000898060296190525,\n",
" 'Connecticut': 0.0005916389731664958,\n",
" 'Delaware': 2.1740967237831155e-05,\n",
" 'District of Columbia': 0.000569204030709808,\n",
" 'Florida': 7.441098676719035e-05,\n",
" 'Guam': 0.0001624030893381615,\n",
" 'Hawaii': 2.6354044452155405e-05,\n",
" 'Idaho': 0.0010328422094282828,\n",
" 'Illinois': 0.00027059063784426246,\n",
" 'Indiana': 3.808868595039972e-05,\n",
" 'Iowa': 0.00028386351044947514,\n",
" 'Kansas': 0.00033797457258202184,\n",
" 'Kentucky': 0.00029640422854047577,\n",
" 'Louisiana': 0.0003446505811545775,\n",
" 'Maine': 8.395350528971605e-05,\n",
" 'Maryland': 0.0010391429532535983,\n",
" 'Massachusetts': 0.00025848147353538844,\n",
" 'Michigan': 0.0018415825748126415,\n",
" 'Minnesota': 4.7428207312129696e-05,\n",
" 'Mississippi': 0.0010291939210410152,\n",
" 'Missouri': 0.0007866767544571429,\n",
" 'Montana': 0.00022008373963125228,\n",
" 'Nebraska': 0.000252896451933831,\n",
" 'Nevada': 4.5528716483467216e-05,\n",
" 'New Hampshire': 0.00013826413183675714,\n",
" 'New Jersey': 0.005348810023114104,\n",
" 'New Mexico': 0.0004152589380814988,\n",
" 'New York': 0.0011330844063905455,\n",
" 'North Carolina': 5.171026856527207e-05,\n",
" 'North Dakota': 0.0002419396036081185,\n",
" 'Ohio': 0.0005485756202291289,\n",
" 'Oklahoma': 0.0003663325351307428,\n",
" 'Oregon': 2.957204035272494e-05,\n",
" 'Pennsylvania': 0.0013292273208941523,\n",
" 'Puerto Rico': 1.9368596885678292e-05,\n",
" 'Rhode Island': 0.00026802882531775713,\n",
" 'South Carolina': 0.0004608273098738372,\n",
" 'South Dakota': 0.0004992767346474372,\n",
" 'Tennessee': 9.879279515508202e-05,\n",
" 'Texas': 0.0002788736212016803,\n",
" 'Utah': 0.0004892900644267518,\n",
" 'Vermont': 0.0008035736622334212,\n",
" 'Virgin Islands': 2.0756865298749132e-05,\n",
" 'Virginia': 1.4701512393884831e-05,\n",
" 'Washington': 0.00014259221679922322,\n",
" 'West Virginia': 0.00028123316420640594,\n",
" 'Wisconsin': 0.0005274779174654732,\n",
" 'Wyoming': 1.1603978586029659e-05,\n",
" 'Uganda': 8.311931995273981e-07,\n",
" 'Ukraine': 1.588412645597683e-05,\n",
" 'United Arab Emirates': 0.00026061449844999606,\n",
" 'Anguilla': 1.8342333564709167e-05,\n",
" 'Bermuda': 6.741287357772619e-05,\n",
" 'British Virgin Islands': 2.189531432540795e-06,\n",
" 'Cayman Islands': 0.00011804547269321272,\n",
" 'Channel Islands': 0.00020187854351237362,\n",
" 'Falkland Islands (Malvinas)': 0.0011110260590320661,\n",
" 'Gibraltar': 0.0009489182099183034,\n",
" 'Isle of Man': 0.0002602646112006958,\n",
" 'Montserrat': 0.0007721579600352,\n",
" 'Turks and Caicos Islands': 0.00010418803030155613,\n",
" 'United Kingdom': 0.0003282761078357052,\n",
" 'Uruguay': 1.5091083130086797e-05,\n",
" 'Uzbekistan': 8.690748043412879e-06,\n",
" 'Venezuela': 4.0066974051167964e-06,\n",
" 'Vietnam': 9.59451065278107e-07,\n",
" 'West Bank and Gaza': 0.0026469110335335735,\n",
" 'Western Sahara': 2.485149350976438e-06,\n",
" 'Zambia': 7.593249524086256e-08,\n",
" 'Zimbabwe': 2.8474272787437325e-07}"
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"val_loss_dict"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[('Ireland', 0.0016420296695800357), ('Quebec', 0.0014828736626624401), ('Pennsylvania', 0.0013292273208941523), ('Denmark', 0.0012591971527977609), ('Turkey', 0.0011840397491182512), ('New York', 0.0011330844063905455), ('Falkland Islands (Malvinas)', 0.0011110260590320661), ('Austria', 0.0010463090882809523), ('Maryland', 0.0010391429532535983), ('Idaho', 0.0010328422094282828)]\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 720x144 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"# losses\n",
"visualize(val_loss_dict, val_info_dict, end=20)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Evaluate Country/Province by Example"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [],
"source": [
"def evaluate(name, val_info_dict=val_info_dict):\n",
" info = val_info_dict[name]\n",
" \n",
" case_split = info[\"Case Split\"]\n",
" fat_split = info[\"Fatality Split\"]\n",
" cases_test = info[\"Cases Actual\"][case_split:]\n",
" fat_test = info[\"Fatalities Actual\"][fat_split:]\n",
" \n",
" case_model = info[\"Case Model\"]\n",
" fat_model = info[\"Fatality Model\"]\n",
" \n",
" print(name)\n",
" print(\"Confirmed Cases:\")\n",
" print(\" Loss Train: \", info[\"Cases Loss Train\"])\n",
" print(\" Loss Test: \", info[\"Cases Loss Test\"])\n",
" display(case_model.evaluate(cases_test))\n",
" print(\"Fatalities:\")\n",
" print(\" Loss Train: \", info[\"Fatality Loss Train\"])\n",
" print(\" Loss Test: \", info[\"Fatality Loss Test\"])\n",
" display(fat_model.evaluate(fat_test))"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Germany\n",
"Confirmed Cases:\n",
" Loss Train: 2.33381364090002e-05\n",
" Loss Test: 8.409149255787794e-05\n",
"Beta: 12.531149570228767\n",
"Gamma: 12.378434406736405\n",
"At t=0: 1.1943743582253332e-08\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Fatalities:\n",
" Loss Train: 1.8398965388451636e-06\n",
" Loss Test: 2.5670051269381823e-05\n",
"Beta: 1.1733371392603076\n",
"Gamma: 1.0014755640373423\n",
"At t=0: 2.3887487164506663e-08\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"evaluate(\"Germany\")"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Spain\n",
"Confirmed Cases:\n",
" Loss Train: 4.052282895712593e-05\n",
" Loss Test: 6.747567701671771e-05\n",
"Beta: 9.307848945525745\n",
"Gamma: 9.13324577721468\n",
"At t=0: 2.138997215432035e-08\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Fatalities:\n",
" Loss Train: 6.432470064248784e-05\n",
" Loss Test: 0.0006616223141886673\n",
"Beta: 0.7732031316965112\n",
"Gamma: 0.5483375624028562\n",
"At t=0: 2.138997215432035e-08\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"evaluate(\"Spain\")"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Hubei\n",
"Confirmed Cases:\n",
" Loss Train: 4.226673712063659e-05\n",
" Loss Test: 1.0548900721327478e-05\n",
"Beta: 17.29318273461616\n",
"Gamma: 17.12789337739622\n",
"At t=0: 7.522873602168756e-06\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Fatalities:\n",
" Loss Train: 8.224879625804729e-06\n",
" Loss Test: 9.56927970034335e-06\n",
"Beta: 99.84970312450345\n",
"Gamma: 99.63317697304076\n",
"At t=0: 2.8803795323619113e-07\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"evaluate(\"Hubei\")"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Italy\n",
"Confirmed Cases:\n",
" Loss Train: 0.0006866289274446315\n",
" Loss Test: 0.0019895647310509694\n",
"Beta: 111.91787349023794\n",
"Gamma: 91.76769924641862\n",
"At t=0: 3.3068304103267285e-08\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Fatalities:\n",
" Loss Train: 5.5144561469408936e-05\n",
" Loss Test: 0.000534957050231585\n",
"Beta: 0.9789425169319386\n",
"Gamma: 0.80319836997655\n",
"At t=0: 1.6534152051633643e-08\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"evaluate(\"Italy\")"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"New York\n",
"Confirmed Cases:\n",
" Loss Train: 0.00021133868939752273\n",
" Loss Test: 0.0011330844063905455\n",
"Beta: -7.061149392747142\n",
"Gamma: -7.299197668607935\n",
"At t=0: 8.89297337387227e-06\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Fatalities:\n",
" Loss Train: 3.9920989201465346e-05\n",
" Loss Test: 0.000540564087435805\n",
"Beta: 1.113548257206262\n",
"Gamma: 0.8549132026101509\n",
"At t=0: 1.0280894073840773e-07\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAZQAAAEWCAYAAABBvWFzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzdeXxU1fn48c+ThYQ1kIUlBExYlR0MKK4VFdCquFFxq7hRVGqttVXb/hRprWvVrytFEdwAEZeioqJFZd+XsEsgIAGEhCXsZHt+f9ybMAyTZEJmMlme96t5MXPvuec+N2PnyT3n3HNEVTHGGGMqKizUARhjjKkZLKEYY4wJCEsoxhhjAsISijHGmICwhGKMMSYgLKEYY4wJCEsopkoQkYMi0iZAdTUTkZkickBE/h2IOqsCEfmViGT6WXakiLwf7JhCqTZcY3VjCcVUKhHZLCJH3ARS9JOoqg1UdZNbZryI/LMCpxkGZAONVPVPPmJIEpGPRSRbRHJEZKWIDHX3JYuIikiERyy5bpx7RORbETm9ArGFXHkSU4DPO1REZlf2eU3lsYRiQuFKN4EU/WwPcP2nAWu05Kd23wO2uuXigN8CO0up71lVbQC0BLYBYwMYqzE1hiUUUyW4dwXtRGQYcDPwF/eu4PMSyp8jIovcO4xFInKOu308cJvH8Zf4OLw3MF5VD6lqvqouU9WvyopRVY8Ak4EepVzHSBH5SETed5vcVopIBxF5VER2ichWEenvUT5RRKa6dz/pInK3x7667h3SXhFZ48aN17Efi0iWiGSIyP1lXYOI1Ae+AhI97xBFJEpEXhKR7e7PSyISVUIdbUVkhojsdu/yPhCRxh77W4nIJ25cu0XkVRE5AxgN9HXPuc8t+4OI3OVx7Al3MSLyf+7vbL+ILBGR88u6RhM6llBMlaKqY4APcO8KVPVK7zIiEgt8CbyMc4fxAvCliMSp6lCv47/zcZr5wGsiMkREWvsbm/tlfCOQXkbRK3HugpoAy4BvcP6/1hIYBfzHo+xEIBNIBK4H/iUiF7v7Hgfauj8DcBJlUSxhwOfACrfei4EHRGRAaYGp6iHgMmC71x3i34CzcZJld6AP8PeSfhXAU27MZwCtgJFuXOHAF8AWINmNbZKqrgWGA/PcczY+uVqfFrkxxQITgI9EJNrPY00ls4RiQuEzEdnn/nx2Csf/Gtigqu+5dxgTgXU4X+T+GAzMAv4fkCEiy0WkdynlH3L/oj4AnAfcWkb9s1T1G1XNBz4CEoCnVTUPmAQki0hjEWnl1vewqh5V1eXAWx71/wZ4UlX3qOpWnARapDeQoKqjVDXX7X96Exji5+/A283AKFXdpapZwBMlXaeqpqvqt6p6zC37AnChu7sPTqL5s3sHeFRVT7nfRFXfV9Xd7uf8byAK6Hiq9ZngsoRiQuFqVW3s/lx9Cscn4vwF7GkLzl/DZVLVvar6iKp2BpoBy3GSnJRwyPPuX9TJwBHK/kLz7I85AmSraoHHe4AG7nXsUdUDJVxHIk5fj+e+IqfhNFsVJeZ9wF/d6zkV3r/TLe62k4hIUxGZJCLbRGQ/8D4Q7+5uBWxxk2mFicifRGSt27S5D4jxOJepYiyhmKqorCmwt+N8oXpqjdNhXr4TqWYDz+N8ecaWUfZn4A/A/4lI3fKey4ftQKyINPTY5nkdO3C+oD33FdkKZHgk5saq2lBVL/fjvL5+v96/09buNl+ecuvopqqNgFtwmsGK4mpdNErOj/MeAup5vG9e9MLtL3kY506tiZvUczzOZaoYSyimKtoJlPZMyjSgg4jcJCIRInID0Amn7b5MIvKMiHRxj20I3AOkq+ruso5V1W9xvmiH+XOuMuraCswFnhKRaBHpBtyJ0wcEzgCAR0WkiYgkAb/3OHwhsF9EHnY778Pdayqt6a7ITiBORGI8tk0E/i4iCSISDzyGc+fhS0PgILBPRFoCf/aKawfwtIjUd6/rXI/zJolIHY/yy4FrRaSeiLRzr9/zPPlAFhAhIo8Bjfy4PhMillBMVTQW6FRSH4v7xX8F8CdgN/AX4Ar3bsMf9YBPgX3AJpy/zK8qR3zP4Ywi8zkKqpxuxGlK2+7G9LibtMDpx9gCZADTcTr6AXCb0K7E6bDOwHnu5i2cJqFSqeo6nASyyf0dJwL/BBYDacBKYKm7zZcngF44dwtfAp/4iKsd8DPOgIMb3N0zgNXALyJS9Fm9COTiJJt3OJ5MwRnM8BXwk/t7OMqJTYCmihFbYMsYY0wg2B2KMcaYgLCEYowxJiAsoRhjjAkISyjGGGMCwtdY8VojPj5ek5OTQx2GMcZUK0uWLMlW1QTv7bU6oSQnJ7N48eJQh2GMMdWKiHjPVAFYk5cxxpgAsYRijDEmICyhGGOMCYha3YfiS15eHpmZmRw9ejTUodQq0dHRJCUlERkZGepQjDGnKKgJRUQGAv8HhANvqerTXvujgHeBM3HmZLpBVTe7+x7FmSiuALhfVb9xt7+NM4/TLlXt4uOcD+HMtZRQjrmdimVmZtKwYUOSk5MpeTZzE0iqyu7du8nMzCQlJSXU4RhjTlHQmrzcldtew1kdrhNwo4h08ip2J7BXVdvhTBL3jHtsJ5yFgjoDA4HX3foAxrvbfJ2zFXApzqR0p+To0aPExcVZMqlEIkJcXJzdFRpTzQWzD6UPzpTgm1Q1F2elukFeZQbhzDAKMAW42F3kaBDOsqHHVDUDZ8nVPgCqOhPYU8I5X8SZebZCM15aMql89js3pvoLZkJpyYlTTWdy8op6xWXcFd5ycNYI9+fYE4jIVcA2VV1RRrlhIrJYRBZnZWX5cx3GGFNj7Du6jxeXvMjmnM0BrzuYCcXXn5zedw4llfHn2OOViNQD/oazKFCpVHWMqqaqampCwkkPelYZn376KSLCunXryiz70ksvcfjw4VM+1/jx4xkxYsQJ28aNG0ePHj3o0aMHderUoWvXrvTo0YNHHnnErzofe+wxvvvuu1OOyRgTHGnZaby96m2yj5S7i7lMwUwomZy4fGkSJy8pWlzGXTI0Bqc5y59jPbUFUoAVIrLZLb9URJqXckyFjf5xI3M3nvihzN2YzegfN1a47okTJ3LeeecxadKkMstWNKH4cvvtt7N8+XKWL19OYmIi33//PcuXL+fpp4+PqygoKCjx+FGjRnHJJZcENCZjTMWtyFpBuITTKc67S7vigplQFgHtRSTFXfJzCDDVq8xU4Db39fXADHVW/JoKDBGRKBFJAdrjLC3qk6quVNWmqpqsqsk4CamXqv4S2Es6UbekGEZMWFacVOZuzGbEhGV0Sypz0bxSHTx4kDlz5jB27NgTEkpBQQEPPfQQXbt2pVu3brzyyiu8/PLLbN++nYsuuoiLLroIgAYNGhQfM2XKFIYOHQrA559/zllnnUXPnj255JJL2LlzZ7lja9CgAY899hhnnXUW8+bNY9SoUfTu3ZsuXbowbNgwihZsGzp0KFOmTAGcKW4ef/xxevXqRdeuXf266zLGBEdaVhodmnSgXmS9gNcdtITi9omMwFnGcy0wWVVXi8got78DnKVe40QkHXgQeMQ9djXOetprgK+B+9ylRRGRicA8oKOIZIqI5xrUleqctvG8elNPRkxYxgvT1zNiwjJevakn57SNr1C9n332GQMHDqRDhw7ExsaydOlSAMaMGUNGRgbLli0jLS2Nm2++mfvvv7/4DuL7778vtd7zzjuP+fPns2zZMoYMGcKzzz5b7tgOHTpEly5dWLBgAeeddx4jRoxg0aJFrFq1iiNHjvDFF76XdY+Pj2fp0qXcc889PP/88+U+rzGm4gq1kFXZq+iW0C0o9Qf1ORRVnQZM89r2mMfro8DgEo59EnjSx/Yb/ThvcnljPVXntI3nlrNa8/KMdO7v167CyQSc5q4HHngAgCFDhjBx4kR69erFd999x/Dhw4mIcD622NjYctWbmZnJDTfcwI4dO8jNzT2lZz7Cw8O57rrrit9///33PPvssxw+fJg9e/bQuXNnrrzyypOOu/baawE488wz+eSTT07ab4wJvoycDA7mHaRrfNeg1G9PylfQ3I3ZvL/gZ+7v1473F/zM2W3jKpRUdu/ezYwZM1i1ahUiQkFBASLCs88+i6r6NbzWs4znsx2///3vefDBB7nqqqv44YcfGDlyZLnji46OJjw8vLjue++9l8WLF9OqVStGjhxZ4rMkUVFRgJOQ8vPzy31eY0zFpWWlAQTtDsXm8qqAoj6TV2/qyYP9OxY3f3l31JfHlClT+O1vf8uWLVvYvHkzW7duJSUlhdmzZ9O/f39Gjx5d/IW8Z4/zOE7Dhg05cOBAcR3NmjVj7dq1FBYW8umnnxZvz8nJoWVLZ/T1O++8Q0UVJY/4+HgOHjxY3GdijKmaVmStoFGdRpzW6LSg1G8JpQLSMnNO6DMp6lNJy8w55TonTpzINddcc8K26667jgkTJnDXXXfRunVrunXrRvfu3ZkwYQIAw4YN47LLLivulH/66ae54oor6NevHy1atCiuZ+TIkQwePJjzzz+f+PiKN801btyYu+++m65du3L11VfTu3fvCtdpjAmetOw0uiZ0JUyC89UvRaNyaqPU1FT1XmBr7dq1nHHGGSGKqHaz370xwXMo7xB9J/Tlnu73cE+PeypUl4gsUdVU7+12h2KMMbXAquxVKBq0/hOwhGKMMbVCUYd8l/iTJmkPGEsoxhhTC6RlpZESk0JMVMUevC6NJRRjjKnhVJW07DS6xQevuQssoRhjTI2XeTCTPUf3BLX/BCyhGGNMjVfUf9I9oXtQz2MJpYqq7tPXByIuY0xgpGWlUTeiLm0btw3qeSyhVMTslyBj5onbMmY62yuoOkxfH4q4jDHll5aVRpf4LkSEBXe2LUsoFdGyF3w09HhSyZjpvG/Zq0LVVuXp65977jl69+5Nt27dePzxxwFnBuJf//rXdO/enS5duvDhhx/6jMsYU/mO5h9l3Z51QZsQ0pNNDlkRKRfA4PFOEkm9ExaPdd6nXFChan1NX9+rV68Tpq+PiIhgz549xMbG8sILL/D999+XOZ1K0fT1IsJbb73Fs88+y7///W+/45o+fTobNmxg4cKFqCpXXXUVM2fOJCsri8TERL788kvAmTMsJibG77iMMcGzbs868jU/6B3yYHcoFZdygZNMZj7r/FvBZAJOc9eQIUOA49PXAwGZvn7AgAF07dqV5557jtWrV5fr+OnTpzN9+nR69uxJr169WLduHRs2bKBr16589913PPzww8yaNYuYmOCNczfGlE/xDMNBHjIMdodScRkznTuTC/7i/JtyfoWSSlWevl5VefTRR/nd73530r4lS5Ywbdo0Hn30Ufr3789jjz3mowZjTGVLy04jsX4iCfUSgn4uu0OpiKI+k8Hjod/fjjd/eXfUl0NVnr5+wIABvP322xw8eBCAbdu2sWvXLrZv3069evW45ZZbeOihh4pXmPSOyxhT+dKy0iqluQssoVTMtqUn9pkU9alsW3rKVVbl6ev79+/PTTfdRN++fenatSvXX389Bw4cYOXKlfTp04cePXrw5JNP8ve//91nXMaYyrXr8C52HNpRaQnFpq+36eurDPvdGxNY/9vyPx744QHev/z9gD7UGJLp60VkoIisF5F0ETnpiTgRiRKRD939C0Qk2WPfo+729SIywGP72yKyS0RWedX1nIisE5E0EflURBoH89qMMaaqW5G9gsiwSM6IrZw/1IKWUEQkHHgNuAzoBNwoIp28it0J7FXVdsCLwDPusZ2AIUBnYCDwulsfwHh3m7dvgS6q2g34CXg0oBdkjDHVTFpWGmfEnkGd8DqVcr5g3qH0AdJVdZOq5gKTgEFeZQYBRb3DU4CLxRmiNAiYpKrHVDUDSHfrQ1VnAnu8T6aq01U13307H0gK9AUZY0x1kV+Yz+rs1ZXWfwLBTSgtga0e7zPdbT7LuMkgB4jz89jS3AF8Vc54jTGmxtiwdwNHC47WmITi64EJ7xEAJZXx51jfJxX5G5APfFDC/mEislhEFmdlZflTpTHGVDvFDzTWkISSCbTyeJ8EbC+pjIhEADE4zVn+HHsSEbkNuAK4WUsYvqaqY1Q1VVVTExKC/6CPMcaEQlp2GrHRsSTWT6y0cwYzoSwC2otIiojUwelkn+pVZipwm/v6emCGmwimAkPcUWApQHtgYWknE5GBwMPAVaparae4DQ8Pp0ePHnTp0oXBgwdXaMbeoUOHMmXKFADuuusu1qxZU2LZH374gblz5xa/Hz16NO++++4pn9sYEzpFDzT6M7tGoAQtobh9IiOAb4C1wGRVXS0io0TkKrfYWCBORNKBB4FH3GNXA5OBNcDXwH2qWgAgIhOBeUBHEckUkTvdul4FGgLfishyERkdrGsLtrp167J8+XJWrVpFnTp1GD36xEspKCg4pXrfeustOnXyHmh3nHdCGT58OL/97W9P6VzGmNDZd3Qfm/dvDvqCWt6C+hyKqk5T1Q6q2lZVn3S3PaaqU93XR1V1sKq2U9U+qrrJ49gn3eM6qupXHttvVNUWqhqpqkmqOtbd3k5VW6lqD/dneDCvrbKcf/75pKen88MPP3DRRRdx00030bVrVwoKCvjzn/9cPJX8f/7zH8CZb2vEiBF06tSJX//61+zatau4rl/96lcUPcj59ddf06tXL7p3787FF1/M5s2bGT16NC+++CI9evRg1qxZjBw5kueffx6A5cuXc/bZZ9OtWzeuueYa9u7dW1znww8/TJ8+fejQoQOzZs2q5N+QMcbbyuyVQOVMCOnJJocsxTMLn2HdnrJXTCyP02NP5+E+D/tVNj8/n6+++oqBA53HbhYuXMiqVatISUlhzJgxxMTEsGjRIo4dO8a5555L//79WbZsGevXr2flypXs3LmTTp06cccdd5xQb1ZWFnfffTczZ84kJSWleBr84cOH06BBAx566CEA/ve//xUf89vf/pZXXnmFCy+8kMcee4wnnniCl156qTjOhQsXMm3aNJ544gm+++67QPyqjDGnaGX2SsIkjM7xnSv1vJZQqqAjR47Qo0cPwLlDufPOO5k7dy59+vQhJSUFcKaST0tLK+4fycnJYcOGDcycOZMbb7yR8PBwEhMT6dev30n1z58/nwsuuKC4rrKmwc/JyWHfvn1ceOGFANx2220MHjy4eP+1114LwJlnnsnmzZsrdvHGmApLy0qjXeN21I+sX6nntYRSCn/vJAKtqA/FW/36x//jUFVeeeUVBgwYcEKZadOmldkJ5+80+P6KiooCnMEERTMhG2NCo1ALSctOY0DygLILB5jNNlxNDRgwgDfeeIO8vDwAfvrpJw4dOsQFF1zApEmTKCgoYMeOHXz//fcnHdu3b19+/PFHMjIygJKnwS8SExNDkyZNivtH3nvvveK7FWNM1bJ5/2YO5B6o9P4TsDuUauuuu+5i8+bN9OrVC1UlISGBzz77jGuuuYYZM2bQtWtXOnTo4POLPyEhgTFjxnDttddSWFhI06ZN+fbbb7nyyiu5/vrr+e9//8srr7xywjHvvPMOw4cP5/Dhw7Rp04Zx48ZV1qUaY8qh6IHGyh7hBTZ9vU1fX4XY796Yihs1bxRfZ3zN7BtnEybBaYQKyfT1xhhjKldaVhpdE7oGLZmUxhKKMcbUEIfzDrNh34ZKnb/LkyUUH2pzM2Co2O/cmIpbvXs1hVoYkg55sIRykujoaHbv3m1fcJVIVdm9ezfR0dGhDsWYam1F1goAusZ3Dcn5bZSXl6SkJDIzM7Gp7StXdHQ0SUm2JpoxFZGWlcZpjU6jcXRoVkC3hOIlMjKy+AlyY4ypLlSVtKw0zkk8J2QxWJOXMcbUABn7M9h9dDc9mvYIWQyWUIwxpgaYs20OAOe2PDdkMVhCMcaYGmDOtjmkxKTQskHLkMVgCcUYY6q5I/lHWPTLIs5NDN3dCVhCMcaYam/xL4vJLczlvJbnhTQOSyjGGFPNzdk+h+jwaFKbnzS9VqWyhGKMMdXc7G2zSW2eSlR4VEjjsIRijDHV2Nb9W9myf0vIm7sgyAlFRAaKyHoRSReRR3zsjxKRD939C0Qk2WPfo+729SIywGP72yKyS0RWedUVKyLfisgG998mwbw2Y4ypCuZsd4YL1+iEIiLhwGvAZUAn4EYR6eRV7E5gr6q2A14EnnGP7QQMAToDA4HX3foAxrvbvD0C/E9V2wP/c98bY0yNNnvbbJIaJNG6YetQhxLUO5Q+QLqqblLVXGASMMirzCDgHff1FOBicRY7HwRMUtVjqpoBpLv1oaozgT0+zudZ1zvA1YG8GGOMqWpyC3JZ+MtCzm15Ls5XZ2gFM6G0BLZ6vM90t/kso6r5QA4Q5+ex3pqp6g63rh1AU1+FRGSYiCwWkcU2AaQxpjpbumspR/KPcH7L80MdChDchOIrXXrPCV9SGX+OPSWqOkZVU1U1NSEhIRBVGmNMSMzOnE1kWCS9m/cOdShAcBNKJtDK430SsL2kMiISAcTgNGf5c6y3nSLSwq2rBbDrlCM3xphqYM72OfRq1ot6kfVCHQoQ3ISyCGgvIikiUgenk32qV5mpwG3u6+uBGeqsbDUVGOKOAksB2gMLyzifZ123Af8NwDUYY0yV9MuhX0jfl15lmrsgiAnF7RMZAXwDrAUmq+pqERklIle5xcYCcSKSDjyIOzJLVVcDk4E1wNfAfapaACAiE4F5QEcRyRSRO926ngYuFZENwKXue2OMqZFmb5sNEPL5uzxJbV7qNjU1VRcvXhzqMIwxptz++P0fWZm9km+v/7bSR3iJyBJVPWmeF3tS3hhjqpm8wjzm75jPeS3PqxLDhYuUuQSwiHQA/gyc5lleVfsFMS5jjDElWLFrBQfzDlaJp+M9+bOm/EfAaOBNoCC44RhjjCnLnO1zCJdwzmpxVqhDOYE/CSVfVd8IeiTGGGP8MmfbHLondKdhnYahDuUE/vShfC4i94pIC3cCxlgRiQ16ZMYYY06SfSSbtXvWcn5S1RkuXMSfO5SiZzv+7LFNgTaBD8cYY0xp5mxzZheuSsOFi5SZUFQ1pTICMcYYU7Y52+YQFx1Hx9iOoQ7lJP6M8ooE7gEucDf9APxHVfOCGJcxxhgvBYUFzN0xlwuTLiRMqt5TH/40eb0BRAKvu+9vdbfdFaygjDHGnGzV7lXkHMupcsOFi/iTUHqraneP9zNEZEWwAjLGGOPbnG1zCJMw+rboG+pQfPLnnqlARNoWvRGRNtjzKMYYU+lmb5tNl/guNI5uHOpQfPLnDuXPwPcisglnnZLTgNuDGpUxxpgT7D26l1XZq7in+z2hDqVE/ozy+p+ItAc64iSUdap6LOiRGWOMKTZv+zwU5dyWVW+4cJESE4qI9FPVGSJyrdeutiKCqn4S5NiMMca4Zm+bTeOoxnSO6xzqUEpU2h3KhcAM4Eof+xSwhGKMMZWgUAuZs30OfRP7Eh4WHupwSlRiQlHVx92Xo1Q1w3Ofu4qiMcaYSrBuzzr2HN1TZYcLF/FnlNfHPrZNCXQgxhhjfCtanfGcxHNCHEnpSutDOR3oDMR49aM0AqKDHZgxxhjH7G2zOSP2DOLrxoc6lFKV1ofSEbgCaMyJ/SgHgLuDGZQxxhjHjoM7WLZrGff2uDfUoZSpxCYvVf2vqt4OXKGqt3v83K+qc/2pXEQGish6EUkXkUd87I8SkQ/d/QtEJNlj36Pu9vUiMqCsOkXkYhFZKiLLRWS2iLTz83dgjDFV1pcZXwJwRZsrQhxJ2fx5sHGZiNyH0/xV3NSlqneUdpCIhAOvAZcCmcAiEZmqqms8it0J7FXVdiIyBHgGuEFEOgFD3HMmAt+5SxFTSp1vAINUda2I3Av8HRjqx/UZY0yVpKpM3TiVXk170aphq1CHUyZ/OuXfA5oDA4AfgSScZq+y9AHSVXWTquYCk4BBXmUGAe+4r6cAF4uIuNsnqeoxd4RZultfaXUqTv8OQAyw3Y8YjTGmylq9ezUZORlc2dbX0xtVjz93KO1UdbCIDFLVd0RkAvCNH8e1BLZ6vM8EvBdALi6jqvkikgPEudvnex3b0n1dUp13AdNE5AiwHzjbV1AiMgwYBtC6dWs/LsMYY0Lj842fUyesDv2T+4c6FL/4c4dStO7JPhHpgvPXf7Ifx4mPbepnmfJuB/gjcLmqJgHjgBd8BaWqY1Q1VVVTExISfAZujDGhlleYx1cZX/GrVr+iUZ1GZR9QBfhzhzJGRJoA/w+YCjQAHvPjuEzAs9EviZOboYrKZIpIBE6y2lPGsSdtF5EEoLuqLnC3fwh87UeMxhhTJc3ZNoe9x/ZWm+Yu8OMORVXfUtW9qvqjqrZR1aaqOtqPuhcB7UUkRUTq4HSyT/UqM5Xja9ZfD8xQVXW3D3FHgaUA7YGFpdS5F+d5maKO+0uBtX7EaIwxVdLUjVNpEtWkSk8G6a20BxsfLO1AVfXZpOSxP19ERuD0t4QDb6vqahEZBSxW1anAWOA9EUnHuTMZ4h67WkQmA2uAfOA+VS1w4zqpTnf73cDHIlKIk2BKHYVmjDFVVc6xHH7c+iODOw4mMiwy1OH4rbQmr4YVrVxVpwHTvLY95vH6KDC4hGOfBJ70p053+6fApxUM2RhjQm76lunkFuZyZZvq09wFpU8O+URlBmKMMcbxxcYvaBPThk5xnUIdSrmU2SkvIuM4eXRWmQ82GmOMKb+t+7eydNdS/tDrDziP5VUf/ozy+sLjdTRwDfbQoDHGBMUXm75AEH6d8utQh1Ju/iwBfML09SIyEfguaBEZY0wtpap8vulzejfvTYsGLUIdTrn582Cjt/aAPWJujDEBtiJrBVsPbK1Wz5548qcP5QDHn1JX4Bfg4SDHZYwxtc7nGz8nOjyaS0+7NNShnBJ/mrwqPHzYGGNM6XILcvl689f0a92P+pH1Qx3OKfGnUx53xcbzcO5QZqnqZ0GNyhhjapmZmTPZn7ufq9peFepQTlmZfSgi8jowHFgJrAKGi8hrwQ7MGGNqk6kbpxJfN56zWnhPyl59+HOHciHQxZ1jCxF5Bye5GGOMCYC9R/cyK3MWN59xMxFhfjUcVUn+jPJaz4mjuloBacEJxxhjap+vN39NvuZX2xIqeaAAACAASURBVNFdRfxJhXHAWhFZ6L7vDcwTkakAqlp9G/yMMaYK+Hzj53Ro0oGOsR1DHUqF+JNQ/Fn7xBhjzCnIyMlgZfZKHkp9KNShVJg/w4Z/FJFmOHcmAAtVdVdwwzLGmNrh842fEyZhXJ5yeahDqTB/Rnn9Bmdxq8HAb4AFInJ9sAMzxpiarlAL+XLTl/Rt0ZeEetV/SXJ/mrz+BvQuuitxl9v9DpgSzMCMMaamW7JzCdsPbef3vX4f6lACwp9RXmFeTVy7/TzOGGNMKT5L/4x6EfXo16pfqEMJCH/uUL4WkW+Aie77G/CxYqIxxhj//XLoF6ZlTOO69tdRL7JeqMMJCH865f/sMfWKAGPc5XaNMcaconfXvIuqcnuX20MdSsCUmFBEpD3wPNAW58n4h1R1W2UFZowxNdW+o/uY8tMULk+5nJYNWoY6nIAprS/kbZzVGq8DlgCvlLdyERkoIutFJF1EHvGxP0pEPnT3LxCRZI99j7rb14vIgLLqFMeTIvKTiKwVkfvLG68xxlSGCesmcCT/CHd0qVkrqZfW5NVQVd90X68XkaXlqVhEwoHXgEuBTGCRiExV1TUexe4E9qpqOxEZAjwD3CAinYAhQGcgEfhORDq4x5RU51CcaWFOV9VCEWlanniNMaYyHMo7xAdrP+CiVhfRrkm7UIcTUKUllGgR6YnTbwJQ1/O9qpaVYPoA6aq6CUBEJgGDAM+EMggY6b6eArwqIuJun6Sqx4AMEUl366OUOu8BblLVQjc+e/jSGFPlTPlpCvtz93NX17tCHUrAlZZQdgAveLz/xeO9AmWNc2sJbPV4nwl4z8tcXEZV80UkB2fusJbAfK9jixoaS6qzLc7dzTVAFnC/qm7wDkpEhgHDAFq3tpWMjTGVJ7cgl3dXv0uf5n3oltAt1OEEXIkJRVUvqmDd4mOb+lmmpO2++nyK6owCjqpqqjsq7W3g/JMKq44BxgCkpqZ6x2OMMUEzdeNUdh3ZxT/O+0eoQwmKYD6gmInTp1EkCdheUhkRiQBigD2lHFtanZnAx+7rT4Gal/6NMdVWQWEB41aNo1NcJ/q26BvqcIIimAllEdBeRFJEpA5OJ/tUrzJTgdvc19cDM9yFvKYCQ9xRYClAe5z5xEqr8zOON8NdCPwUpOsyxphy+3bLt/x84Gfu6noXTldxCMx+CTJmnrgtY6azPQBKTCgicq77b9SpVKyq+cAI4BtgLTBZVVeLyCgRKVpDZSwQ53a6Pwg84h67GpiM09n+NXCfqhaUVKdb19PAdSKyEngKqHk9XsaYaklVeWvlWyQ3Subi1heHLpCWveCjoceTSsZM533LXgGpXtyVfU/eIbJEVc8UkaWqGpizVTGpqam6ePHiUIdhjKnhZmXO4t7/3cuoc0ZxTftrQhtMURJJvRMWj4XB4yHlgnJV4eaHVO/tpY3yyhORcUBLEXnZe6eq2oODxhjjh7dWvkWzes24os0VoQ7FSR6pd8LMZ+GCv5Q7mZSmtD6UK3Calo7iPCnv/WOMMaYMy3YtY+mupQztPJTI8MhQh+PcoSwe6ySTxWNP7lOpgNKGDWcDk0RkraquCNgZjTGmFnlr5Vs0jmrMte2vDXUox5u7ipq5Us4/8X0F+TPKa7eIfCoiu0Rkp4h8LCJJFT6zMcbUcOv3rGdm5kxuPuPm4E5R7+/orW1LT0weKRc477eVa2atEvmTUMbhDM1NxHla/XN3mzHGmFKMXTmWehH1uPH0G4N7In9Hb533wMl3IikXONsDwJ+E0lRVx6lqvvszHqj+ix8bY0wQ/bz/Z77Z8g03dLyBmKiY4J6s6E7jo6Ew48mANmOVhz8JJUtEbhGRcPfnFpxlgI0xxpRg3OpxREgEt3a6tXJO6Dl6K/XOSk8m4F9CuQP4Dc7kkDtwnmivWZP4G2NMAO06vIv/pv+XQe0GkVCvAg065XmyPYijt/xVZkJR1Z9V9SpVTVDVpqp6tapuqYzgjDGmOhqTNoYCLeD2zhVc3tffvhHP0Vv9/na8+auSk0ow5/IyxphaZ1X2Kiavn8yNp99Iq0atyj6gNP72jQR59Ja/SntS3hhjTDkUFBbwj/n/IK5uHPf1uC8wlfrzZLuvUVopF1TJTnljjDF+mPzTZNbsXsNfev+FhnUaBqbSKtA34i+/E4qInC0iM0RkjohcHcygjDGmusk+ks3LS1/m7BZnMzB5YGAqrSJ9I/4qbfr65l6bHgSuAgYCNXO5MWOMOUXPLXqOYwXH+NtZfyt7vZMq8mR7oJV2hzJaRP6fiES77/cBNwE3APuDHpkxxlQTC3YsYFrGNO7ocgfJMcllH1BFnmwPtBITiqpeDSwHvhCRW4EHgEKgHmBNXsYYA+QW5PLP+f8kqUESd3X1c12/KvJke6CV2oeiqp8DA4DGwCfAelV9WVWzKiM4Y4yp6savHs/m/Zv561l/JToiuuwDilSBJ9sDrbQ+lKtEZDYwA1iFs377NSIyUUTaVlaAxhhTVW09sJUxaWO49LRLOT/p/PIdXI1Gb/mrtOdQ/gn0BeoC01S1D/CgiLQHnsRJMMYYUyupKk8teIpwCecvvf9SvoODvC5JqJTW5JWDkzSGALuKNqrqBlX1K5mIyEARWS8i6SLyiI/9USLyobt/gYgke+x71N2+XkQGlKPOV0TkoD/xGWPMqZrx8wxmbZvFvT3upXl9Z1Ds6B83Mndj9gnl5m7MZvSPG088uJqN3vJXaQnlGpwO+Hyc0V3lIiLhwGvAZUAn4EYR6eRV7E5gr6q2A14EnnGP7YSTyDrjDFN+vWi249LqFJFUnP4eY4wJmsN5h3lq4VN0aNKBm8+4uXh7t6QYRkxYVpxU5m7MZsSEZXRL8pq+vpqN3vJXWUsAv1KBuvsA6aq6CUBEJgGDgDUeZQYBI93XU4BXxRnAPQiYpKrHgAwRSXfro6Q63WTzHE7yu6YCcRtjTKneWPEGOw/v5LkLnyMi7PjX6Dlt43n1pp6MmLCMW85qzfsLfubVm3pyTtv4EEZbeYI59UpLYKvH+0x3m88yqpqP08wWV8qxpdU5ApiqqjtKC0pEhonIYhFZnJVlg9WMMeWzYe8G3lvzHte2v5aeTXuetP+ctvHcclZrXp6Rzi1nta41yQSCm1B8PSqqfpYp13YRSQQG48cdlaqOUdVUVU1NSLCFJ40x/ivUQv45/580rNOQP/b6o88yczdm8/6Cn7m/XzveX/DzSX0qNVkwE0om4Dl3cxKwvaQyIhIBxAB7Sjm2pO09gXZAuohsBuq5zWTGGFNhRZ3tH6z9gKW7lvLHM//Imm35J3W2F/WZvHpTTx7s37G4+au2JJVgJpRFQHsRSRGROjid7FO9ykwFbnNfXw/MUFV1tw9xR4GlAO2BhSXVqapfqmpzVU1W1WTgsNvRb4wxFdYtKYZ7P/6I5xc9T79W/Wgq5/nsbE/LzDmhz6SoTyUtMycUYVe6oK2Hoqr5IjIC+AYIB95W1dUiMgpYrKpTgbHAe+7dxB7cZ1vccpNxOvDzgftUtQDAV53BugZjjAFo0zyfekkTOHA4nha5Q7l/4gqfne3DLzz5me9z2sbXmn4UcW4IaqfU1FRdvHhxqMMwxlRhuQW5DP16KJtyNtG/8b94d+YR7u/Xjgf7dwx1aCEjIktUNdV7uy2wZYwxpfjXgn+xMnslt7Z9mC+WFNTKznZ/WUIxxpgSfPTTR3y84WMuS7qZsdMb1NrOdn/ZmvLGmFpt9I8b6ZYUc0I/x9yN2UxPX8h/d/2LcxPPJTn8Wl69KdZnZ3tt6R/xhyUUY0ytVjRdSlEn+9yN2dw36UcatHmV5vWa88wFzxATFXPScbWps91fllCMMbWa93Qp7y3IILnLJ2QeOshbF432mUyMb9aHYoyp9TynS2l/+gzS96fxxDlP0DG29o7kOhWWUIwxtV7RdCmXn72NtYe/4uLE67m8zeWhDqvasYRijKnViqZL+ctVDVh4cAwdGnXnx/ln2QiuU2AJxRhTI/m72FVaZg7/uC6J8elP0DiqMWMG/h+v3ZRaa6ZLCSRLKMaYGsnfxa4GnVmP19b9iX3H9vHSRS8RVzeOc9rG+5xGxZTORnkZY2okfxa7yjyQyZ3f3MmB3AOMuXQMXeK7hDDi6s/uUIwxNVZpi11t2b+FoV8P5VD+Id4c8CbdErqFMNKawRKKMabGKmmxq005m7j969vJLchlbP+xdI7rHOJIawZr8jLG1Eiei12d0zaes9vGMWLCMh4d1JhX1z6EILw94G3aNbGlkwLF7lCMMdVKeUZveS929fBVDXl2xR+IkAjGDRxnySTALKEYY6oVf0dvDb+w7Ql9JquzV/Pymj/RuG59xg0cR0pMSqXGXRtYk5cxplrxZ/SWtxVZKxj+7XBiomIYO2AsLRu0rMSIaw+7QzHGVDuljd7ytnTnUoZNH0ZsdCzjB463ZBJEllCMMVWCv30jRdt9jd7ypKp89NNH3D39bprVb8a4geNoXr950OI3QU4oIjJQRNaLSLqIPOJjf5SIfOjuXyAiyR77HnW3rxeRAWXVKSIfuNtXicjbIhIZzGszxgSWv30jnqO3Slo98VDeIR6Z9Qij5o2id/PejB84nqb1mlbq9dRKqhqUHyAc2Ai0AeoAK4BOXmXuBUa7r4cAH7qvO7nlo4AUt57w0uoELgfE/ZkI3FNWjGeeeaYaY6qOOelZ2nPUdP33N+u056jpOic966Qyb/yQftL2OelZ+sYP6aqqum73Or3ikyu02zvd9D8r/qMFhQWVEnttAixWH9+pweyU7wOkq+omABGZBAwC1niUGQSMdF9PAV4VEXG3T1LVY0CGiKS79VFSnao6rahSEVkIJAXrwowxweHZN3J/v3Y++0Z8zbF1Ttt4+raJ4+OfPuaphU/RqE4j3ur/Fr2b966MsI0rmE1eLYGtHu8z3W0+y6hqPpADxJVybJl1uk1dtwJfV/gKjDEVFui+EV8O5x3mr7P/ysh5I+nZtCeTr5xsySQEgplQxMc29bNMebd7eh2YqaqzfAYlMkxEFovI4qysLF9FjDEBFMi+EV827N3AkC+H8OWmL7m3x72MvmQ08XVtrfdQCGaTVybQyuN9ErC9hDKZIhIBxAB7yji2xDpF5HEgAfhdSUGp6hhgDEBqaqp3MjLGBJi/z434erL91Zt6kpaZU+Kw4E83fMq/FvyL+pH1ebP/m5zV4qygX48pWTATyiKgvYikANtwOt1v8iozFbgNmAdcD8xQVRWRqcAEEXkBSATaAwtx7lB81ikidwEDgItVtTCI12WMKaeK9I34Krvz0E7+vfjffLX5K/o078MzFzxjdyVVQNASiqrmi8gI4Buc0Vlvq+pqERmFM0JgKjAWeM/tdN+DkyBwy03G6cDPB+5T1QIAX3W6pxwNbAHmOf36fKKqo4J1fcYY/3n3jZzdNq7UhxFLcqzgGO+ufpc3V75JQWEB9/W4j7u73k14WHgQojblJc4IsNopNTVVFy9eHOowjKm2Rv+4kW5JMSckh7kbs0nLzCm+4/Ce9df7vT9UlRk/z+C5xc+x7eA2Lm59MX9K/ROtGrYq+2ATcCKyRFVTvbfbk/LGmFPmT4d7aX0j/kjfm87d397NAz88QN2IurzZ/01euuglSyZVkN2h2B2KMRVSlET8najRXznHcnh9+et8uP5D6kfW574e9/Gbjr8hIszmtA21ku5Q7JMxxlSIPx3u5VFQWMCUn6bw6vJX2Z+7n8EdBnNfj/toEt0kQBGbYLGEYow5iT99I57bA9HhfjjvMJ+lf8b7a99n64GtpDZL5ZE+j9AxtmOFr8dUDksoxpiTFPWN+OpI91TSMrvlafbadXgXE9dNZPL6yezP3U+3hG786cw/0a91P9wRm6aasD4U60Mxxid/+kbKcyfjbf2e9by75l2mZUyjoLCAi1tfzG2db6NH0x5BuR4TONaHYowpVwII9MOI4Az/nbN9Du+sfof5O+ZTN6Iuv+nwG2454xZaNbJRW9WdJRRjahF/m7IgcH0j4DzZ/s3mb/g0/VPS96WTUDeBP/T6A4M7DCYmKqbsCky1YAnFmFrE33m1AtE3su/oPqZvmc5XGV+xZOcSFKVTXCeePO9JLku+jMhwWwOvprGEYkwNEOimrFOZqBGclRJn/DyDaRnTmL99PvmaT3KjZO7pfg8DUwaSEpMSoCs2VZElFGNqgEA3ZZWnb+RA7gHmbZ/H15u/ZmbmTI4VHKNF/Rbc2vlWLku+jNNjT7fRWrWEJRRjqjB/7zwqsykrrzCPtKw05u+Yz7zt81iVvYoCLSA2OpZr21/LZSmX0T2hO2FiMzvVNpZQjKnCynPnEaymLFVl476NTgLZMY/FvyzmcP5hwiSMznGduaPLHfRN7EvPpj1tWpRazp5DsedQTAiU90l0f+bKCtScWkfzj7JuzzpW717NyuyVLNyxkKwjzuqmpzU6jbNbnE3fFn1JbZ5qI7RqKXsOxZgqJNB3HqfalJVXmEf63nRW7V7F6uzVrN69mvS96eRrPgDxdeNJbZZK38S+nN3ibBIbJAbul2BqHEsoxgSYP3cf/vZ5FB1bVid6WU1Zqsquw7vYlLOJjJwMNuVsYu2etazfs55jBccAaFSnEZ3jOnN7l9vpHN+ZLnFdaFa/WbB+TaYGsoRijJ/8baby9+4jkHceRefPK8xj24FtTtI4vImt4Zu46csMMnIyOJh3sLh8g8gGdGjSgRs63kCX+C50ietCUsMkG41lKsQSiqn1gpEo/B1xVd47j9TkRjx2TVO+3DCTXwph28Ft7Di4w/n30A52Ht5JoRYWH9+0blNSYlK4os0VtGnchjYxzk983XhLHibgLKGYasffBBCqRFFUtrS7D+87j17JDbjvw//xp8ta0CI2j+wj2WQfzSYrMovVm3fzytpsdh3eRdaRLBRnIM3UnRAmYTSr14zEBomkNkslsUEiSQ2TaBPThpSYFBrWaVixX7Yx5WAJxVRIoL/c/SnnbwKo7EShquQW5nIg9wCzNmby7rKlXHdufd5d+RlH6sYS2zCfnNwc9h3bx8odO2jZ6TD/XHGIfQv2cSDvALSEZ9KOn0sQYqNjia8bT3y9eNo1aUdig0QS6yc6/zZIpGm9pkSG2RQmpmoI6rBhERkI/B8QDrylqk977Y8C3gXOBHYDN6jqZnffo8CdQAFwv6p+U1qdIpICTAJigaXAraqaW1p85R02HIovT3/LhSo277+0vd8Hu9ypDKU9u00sxwqOkVuQy9H8oxwrOMaxgmO8PecnJi/dxFU94hnYLZbDeYc5kn/khJ+MPXuYvXE7SbFhZObsJSkuDJUjHMo7xIG8A+QX5p8Ug6eGkQ1pFNWIxlGNaRzVuPh1Qt0E4uvGE1c3rvh1k+gm9lyHqZJKGjYctIQiIuHAT8ClQCawCLhRVdd4lLkX6Kaqw0VkCHCNqt4gIp2AiUAfIBH4DujgHuazThGZDHyiqpNEZDSwQlXfKC3G8iaUUH95llauvHW9cmMP+raNY87GLO6fuJSXhnTnrJRYCrUQRZm/KYuHPlrBM9d3ITW5CQszsnn0k5X885pO9GrdmEIKKSwspJBClmzZzT++WM2vuzXjy5Xb+ctl7emS2IgCLaBQC4v/XbltL69+/xMXdYzj+592cuf5p9GuaT3yC/Mp0AIKCgso0AI27Mph8pIt9GrdiKVbd3N5t6a0iIkkvzCfvMI88gvzydd8lm/dzdpf9tK+WTTJCdHkFeSRV5hHbkEuuYW55BXmkVeQR/ahQ+w/dpToyAKQfHILS/07o0RRYdEcy4sgvl5D4uo1oqAgio078+ndugUpsfE0iGxAgzoNWLb5CG3j4+mV1IKYqBhi6sTw044CNu4s5L5f2eqDpvoLRULpC4xU1QHu+0cBVPUpjzLfuGXmiUgE8AuQADziWbaonHvYSXUCTwNZQHNVzfc+d0lO5cHGuRuzueeLZ2gYv4qcI7k0bxRNdOSJU0woypHcAnYdOErD6AgOHM0jvmEU0RFhxe3fRb/3o/kF7Dl0jPp1IjiUm0/jehHUCXfKeZbNLSjkwNE8oiLDOJZXQP2ocMLDnHM5/1PyCws5mldARLiQX1BIZAQIUEhhcZmihFHdhEs4EWERxT9aGMb+I0rDqCgOHIWkxg1oUrcekWGRRIZHUiesDpFhkew/oizZfID2CTGk7zrGJacn0Ta+CXXC6xAdHk1URBQ/Z+fy3tztDL/wdHokJbBxZy7Pf72ZJ67sybltEqkXWY/o8Gj+M3PTKS8mZUxNEooHG1sCWz3eZwJnlVTGTQQ5QJy7fb7XsS3d177qjAP2qWq+j/InEJFhwDCA1q1bl++KcNrQ+56WzA9bt9G+aQM6Nmvk1Iuc8C8C6znA+p0H6disAWckxpxQxnOEzZrtB1i7Yz+dWjSiS8vGiMjxsnL8mLTMHFZu20+3ljH0aNWkuJxn+WU/72PZz/vo1TqW3slxJ5UREcIkjPkbd7MgYy9np8RzfvumxfvDJIwwCSt+/cP6Xfz4024u6tCMSzo1J0zCCJfw4nrCJIz0XYcYP2cLF7RryswNu/ndBe3onNikeH+4hBMmYazdfpAXvk3n8s6JfLUqi79e3pkzW8cRHhZenDDCJZylP+fwyJTVDEk9jUmLdvDKjWdyXrumxb+vorus/3jdjf2/Eu7GRnuVu86r3OgdG3n9muOJ4uwWkNKwE2mZOcR1jisuV97FpIypdVQ1KD/AYJw+jqL3twKveJVZDSR5vN+IkxxeA27x2D4WuK6kOnHuatI9trcCVpYV45lnnqnlNSc9S3uOmq7//mad9hw1XeekZ1WZcqE8Z9E+7/fBKPfGD+k+j3vjh/QTtvlbzhhTPsBi9fW972tjIH6AvsA3Hu8fBR71KvMN0Nd9HQFk47TSnFC2qFxJdbrHZAMRvs5d0k95E0oovjz9LReq2AL95W5JwJiqLxQJJQLYBKQAdYAVQGevMvcBo93XQ4DJ7uvObvko9/hNOKO6SqwT+AgY4r4eDdxbVozlTSih+vL0p5x9sRtjKktJCSXYw4YvB15yk8HbqvqkiIxyg5kqItHAe0BPYI+bEDa5x/4NuAPIBx5Q1a9KqtPd3objw4aX4TSZHSstPptt2Bhjyq/SR3lVB5ZQjDGm/EpKKLakmjHGmICwhGKMMSYgLKEYY4wJCEsoxhhjAqJWd8qLSBawxWtzPM4zLdVZTbgGqBnXYddQNdg1BNZpqprgvbFWJxRfRGSxr9EL1UlNuAaoGddh11A12DVUDmvyMsYYExCWUIwxxgSEJZSTjQl1AAFQE64BasZ12DVUDXYNlcD6UIwxxgSE3aEYY4wJCEsoxhhjAsISigcRGSgi60UkXUQeCXU8p0JENovIShFZLiLVYuZLEXlbRHaJyCqPbbEi8q2IbHD/bRLKGMtSwjWMFJFt7mex3J0pu8oSkVYi8r2IrBWR1SLyB3d7tfksSrmGavNZiEi0iCwUkRXuNTzhbk8RkQXu5/ChiNQJdazerA/FJSLhwE/ApThLCC8CblTVNSENrJxEZDOQqqpV5QGoMonIBcBB4F1V7eJuexbYo6pPu8m9iao+HMo4S1PCNYwEDqrq86GMzV8i0gJooapLRaQhsAS4GhhKNfksSrmG31BNPgtx1v2ur6oHRSQSmA38AXgQ+ERVJ4nIaGCFqr4Ryli92R3KcX1wlhHepKq5OGurDApxTLWCqs7EWQ/H0yDgHff1OzhfClVWCddQrajqDlVd6r4+AKwFWlKNPotSrqHacNewOui+jXR/FOgHTHG3V8nPwRLKcS2BrR7vM6lm/yG6FJguIktEZFiog6mAZqq6A5wvCaBpiOM5VSNEJM1tEquyTUXeRCQZZ+G7BVTTz8LrGqAafRYiEi4iy4FdwLfARmCfqua7Rark95MllOPEx7bq2B54rqr2Ai4D7nObYkxovAG0BXoAO4B/hzYc/4hIA+BjnJVS94c6nlPh4xqq1WehqgWq2gNIwmk9OcNXscqNqmyWUI7LBFp5vE8CtocollOmqtvdf3cBn+L8x1gd7XTbw4vaxXeFOJ5yU9Wd7hdDIfAm1eCzcNvsPwY+UNVP3M3V6rPwdQ3V8bMAUNV9wA/A2UBjEYlwd1XJ7ydLKMctAtq7IynqAEOAqSGOqVxEpL7bEYmI1Af6A6tKP6rKmgrc5r6+DfhvCGM5JUVfwq5rqOKfhdsZPBZYq6oveOyqNp9FSddQnT4LEUkQkcbu67rAJTh9Qd8D17vFquTnYKO8PLhDCV8CwoG3VfXJEIdULiLSBueuBCACmFAdrkFEJgK/wpmeeyfwOPAZMBloDfwMDFbVKtvpXcI1/AqniUWBzcDvivoiqiIROQ+YBawECt3Nf8Xpg6gWn0Up13Aj1eSzEJFuOJ3u4Th/9E9W1VHu/78nAbHAMuAWVT0WukhPZgnFGGNMQFiTlzHGmICwhGKMMSYgLKEYY4wJCEsoxhhjAsISijHGmICIKLuIMSYQRKQAZzhrJJCPMzT0JfdhO2OqPUsoxlSeI+50GohIU2ACEIPzzIox1Z41eRkTAu7UOMNwJiwUEUkWkVkistT9OQdARN4TkeJZr0XkAxG5SkQ6u2tmLHcnPGwfqmsxpog92GhMJRGRg6rawGvbXuB04ABQqKpH3eQwUVVTReRC4I+qerWIxADLgfbAi8B8Vf3AnSooXFWPVO4VGXMia/IyJrSKZrmOBF4VkR5AAdABQFV/FJHX3Caya4GPVTVfROYBfxORJJxFlzaEInhjPFmTlzEh4s7NVIAze+8fceYA6w6kAp7Lu74H3AzcDowDUNUJwFXAEeAbEelXeZEb45slFGNCQEQSgNHAq+q0O8cAO9wRX7fiTAxYZDzwAICqrnaPbwNsUtWXcWYD7lZ50RvjmzV5GVN56rqr8BUNG34PKJpi/XXgYxEZjDNN+aGig1R1p4isxZmBqPdjywAAAGlJREFUucgNwC0ikgf8AoyqhPiNKZV1yhtTxYlIPZznV3qpak6o4zGmJNbkZUwVJiKXAOuAVyyZmKrO7lCMMcYEhN2hGGOMCQhLKMb8//bqWAAAAABgkL/1JHaWRMBCKAAshALAQigALAJ9rRye76sF7AAAAABJRU5ErkJggg==\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"evaluate(\"New York\")"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"India\n",
"Confirmed Cases:\n",
" Loss Train: 2.61875346044632e-07\n",
" Loss Test: 2.6424478286460263e-06\n",
"Beta: 0.05846609781595513\n",
"Gamma: -0.022176095730341433\n",
"At t=0: 7.262104630499392e-10\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Fatalities:\n",
" Loss Train: 1.460631700361812e-08\n",
" Loss Test: 5.0982790497263404e-08\n",
"Beta: 1.1839307316481835\n",
"Gamma: 1.0863128287203194\n",
"At t=0: 7.262104630499392e-10\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"evaluate(\"India\")"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"France\n",
"Confirmed Cases:\n",
" Loss Train: 2.986904252118738e-05\n",
" Loss Test: 0.0010104327526954083\n",
"Beta: 0.952407147680054\n",
"Gamma: 0.8341517347357443\n",
"At t=0: 3.0654974021672176e-08\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Fatalities:\n",
" Loss Train: 4.635690407049928e-06\n",
" Loss Test: 9.410958672959614e-05\n",
"Beta: 1.0390972416688997\n",
"Gamma: 0.9022561254969577\n",
"At t=0: 1.5327487010836088e-08\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"None"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"evaluate(\"France\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Submission"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {
"id": "FgSEco90VBdM",
"outputId": "4f78feff-3f74-416e-9b3e-979b31ae8640"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"submission saved to csv.\n"
]
}
],
"source": [
"# submission date range: 02Apr20-14May20\n",
"pd_daterange_submission = pd.date_range(\"02Apr20\", \"14May20\") #TODO get from test dataset: min/max of Date\n",
"length_submission = len(pd_daterange_submission)\n",
"\n",
"def make_submission(val_info_dict=val_info_dict, name=\"submission\"):\n",
" # generate submission frames for all items in val_info_dict\n",
" frames = []\n",
" for attr, item in val_info_dict.items():\n",
" country = item[\"Country\"]\n",
" province = item[\"Province\"]\n",
" case_length = item[\"Case length\"]\n",
" fat_length = item[\"Fatality length\"]\n",
" case_model = item[\"Case Model\"]\n",
" fat_model = item[\"Fatality Model\"]\n",
"\n",
" if(type(province)==float):\n",
" pop = get_population(country)\n",
" else:\n",
" pop = get_population(country, province)\n",
" \n",
" case_preds = pop * case_model.predict(case_length + length_submission)[case_length:]\n",
" fat_preds = pop * fat_model.predict(fat_length + length_submission)[fat_length:]\n",
"\n",
" frames.append(pd.DataFrame({\n",
" \"Country_Region\": country,\n",
" \"Province_State\": province,\n",
" \"Date\": pd_daterange_submission,\n",
" \"ConfirmedCases\": case_preds,\n",
" \"Fatalities\": fat_preds\n",
" })\n",
" )\n",
" \n",
" # concat sub frames and prepare for mergeing with test to get ForecastId\n",
" submission_data = pd.concat(frames)\n",
" submission = test.copy()\n",
"\n",
" index = [\"id\", \"Date\"]\n",
" submission[\"id\"] = submission[\"Country_Region\"].astype(str) + \"_\" + submission[\"Province_State\"].astype(str)\n",
" submission = submission[[\"id\", \"Date\", \"ForecastId\"]].set_index(index)\n",
"\n",
" submission_data[\"id\"] = submission_data[\"Country_Region\"].astype(str) + \"_\" + submission_data[\"Province_State\"].astype(str)\n",
" submission_data = submission_data[[\"id\", \"Date\", \"ConfirmedCases\", \"Fatalities\"]].set_index(index)\n",
"\n",
" # merge w/ ForecastId and extract submission columns\n",
" submission = submission.join(submission_data)\n",
" submission = submission[[\"ForecastId\", \"ConfirmedCases\", \"Fatalities\"]]\n",
"\n",
" # fillna (China)\n",
" submission = submission.fillna(1)\n",
" \n",
" # write to csv\n",
" submission.to_csv(name + \".csv\", index=False)\n",
"\n",
" print(\"submission saved to csv.\")\n",
" \n",
"make_submission()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# [WIP] Temporal SIR-Model\n",
"Add temporal variability to SIR-Model's parameters"
]
},
{
"cell_type": "code",
"execution_count": 25,
"metadata": {},
"outputs": [],
"source": [
"class SIRT:\n",
" def __init__(self, gamma=0, a=0, b=0, c=0, d=0, fix_gamma=False):\n",
" self.gamma = gamma\n",
" self.a = a\n",
" self.b = b\n",
" self.c = c\n",
" self.d = d\n",
" self.infected_t0 = 0\n",
" self.fitted_on = np.array([])\n",
" self.fix_gamma = fix_gamma\n",
" self.fitted = False\n",
" \n",
" def ode(self, y, timestep, c, d, gamma):\n",
" '''Defines the ODE that governs the SIRs behaviour'''\n",
" beta = c * timestep + d\n",
" \n",
" dSdt = -beta * y[0] * y[1]\n",
" dRdt = gamma * y[1]\n",
" dIdt = -(dSdt + dRdt)\n",
" return dSdt, dIdt, dRdt\n",
" \n",
" def solve_ode(self, x, c, d, gamma):\n",
" '''Solves the resulting ODE to get predictions for each time step'''\n",
" return np.cumsum(integrate.odeint(self.ode, (1-self.infected_t0, self.infected_t0, 0.0), x, args=(c, d, gamma))[:,1])\n",
" \n",
" def solve_ode_fixed(self, x, beta):\n",
" '''Solves the resulting ODE to get predictions for each time step'''\n",
" return np.cumsum(integrate.odeint(self.ode, (1-self.infected_t0, self.infected_t0, 0.0), x, args=(beta, self.gamma))[:,1])\n",
" \n",
" def describe(self):\n",
" assert self.fitted, \"You need to fit the model before describing it!\"\n",
" print(\"c: \", self.c)\n",
" print(\"d: \", self.d)\n",
" print(\"Gamma: \", self.gamma)\n",
" print(\"Infected at t=0: \", self.infected_t0)\n",
" \n",
" plt.plot(range(1,len(self.fitted_on)+1), self.fitted_on, \"x\", label='Actual')\n",
" plt.plot(range(1,len(self.fitted_on)+1), self.predict(len(self.fitted_on)), label='Prediction')\n",
" plt.title(\"Fit of SIR model to global infected cases\")\n",
" plt.ylabel(\"Population infected\")\n",
" plt.xlabel(\"Days\")\n",
" plt.legend()\n",
" plt.show()\n",
" \n",
" def fit(self, y):\n",
" '''Fits the parameters to the data, assuming the first data point is the start of the outbreak'''\n",
" self.infected_t0 = y[0]\n",
" x = np.array(range(1,len(y)+1), dtype=float)\n",
" self.fitted_on = y\n",
" if(self.fix_gamma):\n",
" popt, _ = optimize.curve_fit(self.solve_ode_fixed, x, y)\n",
" self.beta = popt[0]\n",
" else:\n",
" popt, _ = optimize.curve_fit(self.solve_ode, x, y)\n",
" self.c = popt[0]\n",
" self.d = popt[1]\n",
" self.gamma = popt[2]\n",
" self.fitted = True\n",
" \n",
" def predict(self ,length):\n",
" '''Returns the predicted cumulated cases at each time step, assuming outbreak starts at t=0'''\n",
" #assert self.fitted, \"You need to fit the model before predicting!\"\n",
" return self.solve_ode(range(1, length+1), self.c, self.d, self.gamma)"
]
},
{
"cell_type": "code",
"execution_count": 26,
"metadata": {},
"outputs": [],
"source": [
"measures = pd.read_csv(\"../input/covid19-containment-and-mitigation-measures/COVID 19 Containment measures data.csv\")\n",
"measures[\"Keywords\"].fillna(value=\"-\", inplace=True)\n",
"measures[\"Country\"] = measures[\"Country\"].str.replace('South Korea', 'Korea, South', regex=True)\n",
"measures[\"Country\"] = measures[\"Country\"].str.replace('US:Georgia', 'US', regex=True)\n",
"measures[\"Country\"] = measures[\"Country\"].str.replace('US: Illinois', 'US', regex=True)\n",
"measures[\"Country\"] = measures[\"Country\"].str.replace('US:Maryland', 'US', regex=True)\n",
"\n",
"measures = measures[measures[\"Country\"] != \"Vatican City\"]\n",
"measures = measures[measures[\"Country\"] != \"Hong Kong\"]\n",
"\n",
"def get_measures(measure_name):\n",
" \n",
" took_measure = measures[measures[\"Keywords\"].str.contains(\"distancing\")]\n",
" output = pd.DataFrame(data=0,\n",
" columns=train['Country_Region'].unique(),\n",
" index=pd.date_range(\"02.01.2020\", \"03.01.2020\"))\n",
" \n",
" print(took_measure)\n",
" \n",
" for index, row in took_measure.iterrows():\n",
" output[row[\"Country\"]][pd.to_datetime(row[\"Date Start\"]):] = 1\n",
" return output\n",
" \n",
"#get_measures(\"distancing\")[\"Italy\"]"
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"c: 45.99821649993995\n",
"d: 2081.254522256052\n",
"Gamma: 7.184669771794915\n",
"Infected at t=0: 2.138997215432035e-08\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"model = SIRT()\n",
"c, _, case_split, _, case_length, _ = get_country_data(\"Spain\")\n",
"model.fit(c[:case_split])\n",
"model.describe()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# [WIP] Modeling SIR Parameters\n",
"Predicting SIR parameters from Country/Province Metadata"
]
},
{
"cell_type": "code",
"execution_count": 28,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(array([2.13899722e-08, 2.13899722e-08, 2.13899722e-08, 2.13899722e-08,\n",
" 2.13899722e-08, 2.13899722e-08, 2.13899722e-08, 2.13899722e-08,\n",
" 4.27799443e-08, 4.27799443e-08, 4.27799443e-08, 4.27799443e-08,\n",
" 4.27799443e-08, 4.27799443e-08, 4.27799443e-08, 4.27799443e-08,\n",
" 4.27799443e-08, 4.27799443e-08, 4.27799443e-08, 4.27799443e-08,\n",
" 4.27799443e-08, 4.27799443e-08, 4.27799443e-08, 4.27799443e-08,\n",
" 1.28339833e-07, 2.78069638e-07, 3.20849582e-07, 6.84479109e-07,\n",
" 9.62548747e-07, 1.79675766e-06, 2.56679666e-06, 3.52934541e-06,\n",
" 4.74857382e-06, 5.54000279e-06, 8.55598886e-06, 1.06949861e-05,\n",
" 1.43954513e-05, 2.29514401e-05, 3.62560028e-05, 4.87049666e-05,\n",
" 4.87049666e-05, 1.11912334e-04, 1.36703312e-04, 1.66799003e-04,\n",
" 2.12659103e-04, 2.51289393e-04, 2.97534513e-04, 3.84228070e-04,\n",
" 4.36569332e-04, 5.42749153e-04, 6.15346719e-04, 7.51558062e-04,\n",
" 8.53139039e-04, 1.05912447e-03, 1.23604093e-03, 1.40572758e-03,\n",
" 1.56649461e-03, 1.71355067e-03, 1.88137639e-03, 2.05179030e-03,\n",
" 2.22708112e-03, 2.39706723e-03, 2.54966329e-03, 2.69873001e-03,\n",
" 2.81590427e-03, 2.92347444e-03, 3.03613543e-03, 3.17042167e-03,\n",
" 3.27741431e-03, 3.38545506e-03, 3.48714299e-03, 3.56851044e-03,\n",
" 3.63841287e-03, 3.69064719e-03]),\n",
" array([2.13899722e-08, 4.27799443e-08, 6.41699165e-08, 1.06949861e-07,\n",
" 2.13899722e-07, 3.63629527e-07, 5.98919220e-07, 7.48649025e-07,\n",
" 1.15505850e-06, 1.17644847e-06, 2.84486630e-06, 4.17104457e-06,\n",
" 6.18170195e-06, 7.31537048e-06, 1.14008552e-05, 1.33259527e-05,\n",
" 1.77536769e-05, 2.23097410e-05, 2.94112117e-05, 3.79030307e-05,\n",
" 4.94322256e-05, 6.00630418e-05, 7.80092284e-05, 9.33672285e-05,\n",
" 1.09901677e-04, 1.27954813e-04, 1.45515981e-04, 1.65045025e-04,\n",
" 1.81044724e-04, 2.00787669e-04, 2.21343432e-04, 2.39524908e-04,\n",
" 2.55545997e-04, 2.70390638e-04, 2.85363619e-04, 3.00422159e-04,\n",
" 3.16400468e-04, 3.30410900e-04, 3.43972142e-04, 3.55201878e-04,\n",
" 3.68100031e-04, 3.79800346e-04, 3.86217337e-04]),\n",
" 74,\n",
" 43,\n",
" 74,\n",
" 43)"
]
},
"execution_count": 28,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"get_country_data(\"Spain\")"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {},
"outputs": [],
"source": [
"for attr, item in val_info_dict.items():\n",
" country = item[\"Country\"]\n",
" province = item[\"Province\"]\n",
" case_length = item[\"Case length\"]\n",
" fat_length = item[\"Fatality length\"]\n",
" case_model = item[\"Case Model\"]\n",
" fat_model = item[\"Fatality Model\"]"
]
}
],
"metadata": {
"colab": {
"collapsed_sections": [],
"name": "LeoCorona",
"provenance": [],
"toc_visible": true
},
"kernelspec": {
"display_name": "Python 3",
"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": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment