Hi Kyle,
Thank you for the hint, really a good idea :-)
--
yegle
http://about.me/yegle
On Sep 12, 2013, at 10:14 PM, Kyle Kastner wrote:
> You can do some regexes for the tree in string form directly to modify the
> output, then write out the modified dotfile. As an example (from some code I
You can do some regexes for the tree in string form directly to modify the
output, then write out the modified dotfile. As an example (from some code
I recently did):
def customize_graphviz(string, true="Output = True" , false="Output =
False"):
import re
string = re.sub(r'entropy = [0-9]+
Thank you Olivier, this does solves my problem. Although not very perfect.
Now the node was marked with "humidity=high <= 0.5000". It'll be great if I can
remove the "<=0.5000" part.
--
yegle
http://about.me/yegle
On Sep 12, 2013, at 3:16 AM, Olivier Grisel wrote:
> You can pass a feature_nam
You can pass a feature_names param to the tree export method. To get the
vectorized feature names, use the vectorizer.get_feature_names() method.
--
How ServiceNow helps IT people transform IT departments:
1. Consolidate l
Dear Yegle,
1) What does your data represent? Are your features numbers or concepts?
In the first case, you should try to build your estimator without
encoding anything. In the second case, it might also not be necessary
to one-hot encode your categorical features. Try with and without
encoding a
Hi list,
I'm a beginner in Machine Learning and trying to write a classifier using
training set containing categorical values.
From the document [1] I learned that I need to encode (vectorize) my
categorical features in order to be learned by the classifier. So I uses
`DictVectorizer` to do th
