I just wanted to address why I wanted to look at the code. I am trying to learn about ensemble methods. In your documentation, you usually give the references which allows me to read the original paper. When there is something in the paper that I don't understand, I like to try to look at the source code. This helps me to understand how the algorithm should work.
In terms of documentation, I think the documentation is very extensive. However, as someone who is new to both machine learning and scikit - learn, I do get confused sometimes. Here's an example: Both of the links below refer to the same (I think) Decision Tree Classifier. But on each page, the model is called in different ways. On this page (http://scikit-learn.org/0.13/modules/tree.html), the example shows from sklearn import tree clf = tree.DecisionTreeClassifier() On this page (http://scikit-learn.org/0.13/modules/generated/sklearn.tree.DecisionTreeClassifier.html), the example shows from sklearn.tree import DecisionTreeClassifier clf = DecisionTreeClassifier(random_state=0) Now that I have used scikit - learn more, I know that these are equivalent, but it was very confusing when I was first trying to understand the underlying structure. It would be helpful if the examples used one consistent way to use the model. Additionally, in this model, you can input a vector of sample weights. It isn't clear to me how this model uses the sample weights. It may be a lack of knowledge on my part. However, several of the references on that page refer to ensemble methods that use sample weights. Anne Dwyer On Tue, Apr 30, 2013 at 9:34 AM, Jaques Grobler <[email protected]> wrote: > Perhaps this is a good place to mention this proposed enhancement > https://github.com/scikit-learn/scikit-learn/issues/1680 > > Perhaps having a direct link to the code would be useful in this case? > > > > 2013/4/30 Peter Prettenhofer <[email protected]> >> >> You can find the source code here (just navigate through the package >> structure): >> >> https://github.com/scikit-learn/scikit-learn/tree/master/sklearn >> >> In case you use IPython to prototype your solution you can simply type two >> question marks to read the source code:: >> >> DecisionTreeClassifier?? >> >> >> >> 2013/4/30 Anne Dwyer <[email protected]> >>> >>> Sometimes I like to look at how the code is written to answer my own >>> questions about how the model works. How can I find the code for a >>> particular model? >>> >>> Anne Dwyer >>> >>> >>> ------------------------------------------------------------------------------ >>> Introducing AppDynamics Lite, a free troubleshooting tool for Java/.NET >>> Get 100% visibility into your production application - at no cost. >>> Code-level diagnostics for performance bottlenecks with <2% overhead >>> Download for free and get started troubleshooting in minutes. >>> http://p.sf.net/sfu/appdyn_d2d_ap1 >>> _______________________________________________ >>> Scikit-learn-general mailing list >>> [email protected] >>> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general >>> >> >> >> >> -- >> Peter Prettenhofer >> >> >> ------------------------------------------------------------------------------ >> Introducing AppDynamics Lite, a free troubleshooting tool for Java/.NET >> Get 100% visibility into your production application - at no cost. >> Code-level diagnostics for performance bottlenecks with <2% overhead >> Download for free and get started troubleshooting in minutes. >> http://p.sf.net/sfu/appdyn_d2d_ap1 >> _______________________________________________ >> Scikit-learn-general mailing list >> [email protected] >> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general >> > > > ------------------------------------------------------------------------------ > Introducing AppDynamics Lite, a free troubleshooting tool for Java/.NET > Get 100% visibility into your production application - at no cost. > Code-level diagnostics for performance bottlenecks with <2% overhead > Download for free and get started troubleshooting in minutes. > http://p.sf.net/sfu/appdyn_d2d_ap1 > _______________________________________________ > Scikit-learn-general mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/scikit-learn-general > ------------------------------------------------------------------------------ Introducing AppDynamics Lite, a free troubleshooting tool for Java/.NET Get 100% visibility into your production application - at no cost. Code-level diagnostics for performance bottlenecks with <2% overhead Download for free and get started troubleshooting in minutes. http://p.sf.net/sfu/appdyn_d2d_ap1 _______________________________________________ Scikit-learn-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
