JLFDataScience

from sklearn.metrics import accuracy_score, auc, confusion_matrix, f1_score, precision_score, recall_score, roc_curve

def metrics_models(y_true, y_pred):
    from sklearn.metrics import accuracy_score, auc, confusion_matrix, f1_score, precision_score, recall_score, roc_curve

    # Obtención de matriz de confusión
    confusion_matrix = confusion_matrix(y_true, y_pred)

    print("La matriz de confusión es ")
    print(confusion_matrix)

JLFDataScience

data_model_1 = pd.concat([diagnostic[categorical_1], 
                          pd.get_dummies(diagnostic['a1_g1_b'], 
                          prefix = 'pclass')], axis = 1)
data_model_2 = pd.concat([diagnostic[categorical_2], 
                          pd.get_dummies(diagnostic['a1_g2'], 
                          prefix = 'pclass')], axis = 1)

JLFDataScience

#Information Value Function
def calculateIV(data, features, target):
    result = pd.DataFrame(index = ['IV'], columns = features)
    result = result.fillna(0)
    var_target = array(data[target])
    
    for cat in features:
        var_values = array(data[cat])
        var_levels = unique(var_values)

JLFDataScience

def get_WoE(data, var, target):
    crosstab = pd.crosstab(data[target], data[var])
    
    print("Obteniendo el Woe para la variable", var, ":")
    
    for col in crosstab.columns:
        if crosstab[col][1] == 0:
            print("  El WoE para", col, "[", sum(crosstab[col]), "] es infinito")
        else:
            print("  El WoE para", col, "[", sum(crosstab[col]), "] es", np.log(float(crosstab[col][0]) / 

JLFDataScience

#We generate a function with the above tests to calculate the Beta of each record
def BetaCalculated(ticker_symbol, ref_index, source_web, sdate, edate):
    try:
        df_stock = web.DataReader(ticker_symbol,source_web,sdate,edate)
        df_index = web.DataReader(ref_index,source_web,sdate,edate)
        #df_stock = pdr.get_data_yahoo(ticker_symbol, start=sdate, end=edate)
        #df_index = pdr.get_data_yahoo(ref_index, start=sdate, end=edate)
        # create a time-series of monthly data points
        df_stock = df_stock.resample('M').last() #We group by Month, i.e. one record per month of each year
        df_index = df_index.resample('M').last() #We group by Month

JLFDataScience

print(edate)
numerator = covariance(df['stock_returns'],df['index_returns'])
print("COVARIANCE(stock, benchmark) = COVARIANCE("+ticker_symbol+", "+ref_index +") = " +str(numerator))
denominator = covariance(df['index_returns'],df['index_returns'])
print("VARIANCE(benchmark) = COVARIANCE(benchmark, benchmark) = COVARIANCE("+ref_index+", "+ref_index +") = " +str(denominator))

# BETA = Covariance (stock,index) / Variance (Index) = Covariance (stock,index) / Covariance (stock,stock)
print("BETA = COVARIANCE(stock, benchmark) / VARIANCE(benchmark) = " + str(numerator) + " / " + str(
                            denominator) + " = " +str(covariance(df['stock_returns'],df['index_returns'])/covariance(
                            df['index_returns'],df['index_returns'])))

JLFDataScience

#We define a function to calculate the covariance
def covariance(a, b):
    if len(a) != len(b):
        return
    a_mean = np.mean(a)
    b_mean = np.mean(b)
    sum = 0
    for i in range(0, len(a)):
        sum += ((a[i] - a_mean) * (b[i] - b_mean))
    return sum/(len(a)-1)

JLFDataScience

#Calculating returns
# -*- coding: utf-8 -*-
"""
reference: http://gouthamanbalaraman.com/blog/calculating-stock-beta.html
"""
df_stock['returns'] = df_stock['Adj Close']/ df_stock['Adj Close'].shift(1) -1
df_stock = df_stock.dropna()
df_index['returns'] = df_index['Adj Close']/ df_index['Adj Close'].shift(1) -1
df_index = df_index.dropna()

JLFDataScience

m = folium.Map(location=[40,-4], zoom_start=6, width=700, height=500, control_scale=True, tiles='CartoDB Positron') 

folium.Choropleth(
    geo_data=geojson_counties,
    name='Comunidades Autónomas',
    data=df_casos,
    columns=['CCAA', 'Casos'],
    key_on='feature.properties.texto',   #'features.properties.comunidade_autonoma'
    fill_color='YlGn',
    fill_opacity=0.5,

JLFDataScience

m = folium.Map(location=[40,-4], zoom_start=6, width=700, height=500, control_scale=True, tiles='CartoDB Positron') 

folium.Choropleth(
    geo_data=geojson_counties,
    name='Comunidades Autónomas',
    data=df_casos,
    columns=['CCAA', 'Casos'],
    key_on='feature.properties.texto',   #'features.properties.comunidade_autonoma'
    fill_color='YlGn',
    fill_opacity=0.5,