Orange Data Mining and Logistic Regression

Question

For logistic regression in R can I get the list of significant factors and other details as we get in R ? In R when we run glm function we get all the details like coefficient and whether the independent variable is significant or not. Can we do similar thing in Orange?

Answer 1

A statistical classification method that fits data to a logistic function is known as logistic regression. Orange adds features to the approach such as stepwise variable selection and handling of constant variables and singularities.
The code below will help you understand a few things:

import Orange
data = Orange.data.Table("titanic")
log_reg = Orange.classification.logreg.LogRegLearner(data)

# compute classification accuracy

correct = 0.0
for feature in titanic:

    if log_reg(feature) == feature.getclass():

        correct += 1

print "Classification accuracy:", correct / len(data)
print Orange.classification.logreg.dump(log_reg)

Output

Classification accuracy: 0.778282598819

class attribute = survived
class values = <no, yes>

      Feature       beta  st. error     wald Z          P OR=exp(beta)

    Intercept      -1.23       0.08     -15.15      -0.00
 status=first       0.86       0.16       5.39       0.00 2.35e0
status=second      -0.16       0.18      -0.91       0.36 8.51e-1
 status=third      -0.92       0.15      -6.12       0.00 3.98e-1
    age=child       1.06       0.25       4.30       0.00 2.89e0
   sex=female       2.42       0.14      17.04       0.00 1.12e1

This is taken from the orange 2.7 documentation. Hope this helps to answer your question.