rloliveirajr / sklearn_transformers

Implements several sklearn transformers to several applications.

Trans4mers

Trans4mers is a sklearn compatible python package that implements some methods of data preprocessing. Currently the methods implemented are:

• Imputation:
  • KNN Regression: This method imput values for missing values based on KNN Regression.
• Feature Extraction: I'm not sure if the above methods are really feature extractors.
  • Hyberbolic Location Fingerprint: This method is used on Indoor Wifi Localization when there are heterogeneous devices environment. For more details: M. Kjaergaard and C. Munk, "Hyperbolic Location Fingerprint: A calibration-free solution for handling differences in signal strength", in PerCom, pp. 110-116, 2008
  • Relative Location Fingerprint: This method is just a variation of the above. Instead of use Hyberbolic tranformation it just take the relation betwee two variables.
• Normalization:
  • Standard Normalization: This method normalize each row of a dataset based on the mean and standard variation among the variables of the row.

Instaling

First clone the repository to your machine:

git clone [e-mail]:rloliveirajr/sklearn_transformers.git

Then, install the package using pip or easy_install:

pip setup.py install

Using

...

Contributions

Any contribution is welcome!