Azure ML Scripting with Python

Evangelist at Microsoft
@timmyreilly
Timmyreilly.com

Free Machine Learning Platform
https://studio.azureml.net/

Full ML Documentation:
https://azure.microsoft.com/en-us/documentation/services/machine-learning/

Python ML Documentation:
https://azure.microsoft.com/en-us/documentation/articles/machine-learning-execute-python-scripts/

Gallery of Projects:
https://gallery.cortanaanalytics.com/

Gallery of Projects tagged Python:
https://gallery.cortanaanalytics.com/browse/?skip=0&orderby=trending%20desc&tags=%5B%22Python%22%5D

Breast Cancer Correlation Matrix:
https://gallery.cortanaanalytics.com/Experiment/Correlation-Matrix-2

def azureml_main(dataframe1 = None):
    import numpy as np
    import pandas as pd
    import matplotlib
    matplotlib.use("agg")  
    import matplotlib.pyplot as plt
    cm=dataframe1.corr()
    
    fig=plt.figure()
    plt.imshow(cm,interpolation='nearest')
    plt.xticks(list(range(0,len(cm.columns))),list(cm.columns.values), rotation=45)
    plt.yticks(list(range(0,len(cm.columns))),list(cm.columns.values))
    
    plt.colorbar()


    fig.savefig("CorrelationPlot.png")

    return pd.DataFrame(cm), 

Feature Importance with sklearn
https://gallery.cortanaanalytics.com/Experiment/Feature-Importance-with-sklearn-2

import numpy as np
import matplotlib
matplotlib.use("agg")
import matplotlib.pyplot as plt
import pandas as pd


from sklearn.ensemble import ExtraTreesClassifier


def azureml_main(iris_data):
    X = iris_data[["sepal-length","sepal-width","petal-length","petal-width"]]
    Y = iris_data["Class"]
    
    etc = ExtraTreesClassifier(n_estimators=250, random_state=0)
    etc.fit(X,Y)
    
    feat_imp = etc.feature_importances_
    std = np.std([tree.feature_importances_ for tree in etc.estimators_],\
                 axis=0)
    indices = np.argsort(feat_imp)[::-1]
    length_fp =  len(indices)

    fig = plt.figure()
    plt.title("Feature importances")
    plt.bar(range(length_fp), feat_imp[indices],\
            color="r", yerr=std[indices], align="center")
    plt.xticks(range(length_fp), indices)
    plt.xlim([-1, length_fp])
    plt.xlabel("Feature")
    plt.ylabel("Importance")
    fig.savefig("Feature_Importance.png")
    return iris_data

Beer Forecasting Example:
https://gallery.cortanaanalytics.com/Experiment/Forecasting-Beer-Consumption-2

Leave a Reply

Your email address will not be published. Required fields are marked *