In [ ]:

    

from azureml import Workspace
ws = Workspace()
experiment = ws.experiments['6aa07b1af33b4b2bb32adc11b6e40bdc.f-id.299df9aa87674e0da3349774b2df8879']
ds = experiment.get_intermediate_dataset(
    node_id='e77d289d-9cde-4a3f-a13d-42425508c22b-314',
    port_name='Results dataset',
    data_type_id='GenericCSV'
)
Income = ds.to_dataframe()


    

In [ ]:

    

Income.dtypes


    

In [ ]:

    

## Plot categorical variables as bar plots
def income_barplot(df):
    import numpy as np
    import matplotlib.pyplot as plt
    
    cols = df.columns.tolist()[:-1]
    for col in cols:
        if(df.ix[:, col].dtype not in [np.int64, np.int32, np.float64]):
            temp1 = df.ix[df['income'] == ' <=50K', col].value_counts()
            temp0 = df.ix[df['income'] == ' >50K', col].value_counts() 
            
            ylim = [0, max(max(temp1), max(temp0))]
            fig = plt.figure(figsize = (12,6))
            fig.clf()
            ax1 = fig.add_subplot(1, 2, 1)
            ax0 = fig.add_subplot(1, 2, 2) 
            temp1.plot(kind = 'bar', ax = ax1, ylim = ylim)
            ax1.set_title('Values of ' + col + '\n for income <= 50K')
            temp0.plot(kind = 'bar', ax = ax0, ylim = ylim)
            ax0.set_title('Values of ' + col + '\n for income > 50K')
    return('Done')


    

In [ ]:

    

income_barplot(Income)


    

In [ ]:

    

## Plot categorical variables as box plots
def income_boxplot(df):
    import numpy as np
    import matplotlib.pyplot as plt
    
    cols = df.columns.tolist()[:-1]
    for col in cols:
        if(df[col].dtype in [np.int64, np.int32, np.float64]):                  
            fig = plt.figure(figsize = (6,6))
            fig.clf()
            ax = fig.gca() 
            df.boxplot(column = [col], ax = ax, by = [' income'])          
    return('Done')


    

In [ ]:

    

income_boxplot(Income)