search

pixelogik / NearPy

Python framework for fast (approximated) nearest neighbour search in large, high-dimensional data sets using different locality-sensitive hashes.
MIT License
759 stars 152 forks source link

Improvement on README about role of RandomBinaryProjections #8

Closed linearregression closed 9 years ago

linearregression commented 10 years ago

The information about role of RandomBinaryProjections is sparse. Please improve that?

The code for experiment is clearer about that they are doing then the example usage. Suggest improve example usage. Also have a folder 'example' containing self contained samples.

pixelogik commented 10 years ago

Well the documentation in the RandomBinaryProjections class says very clear, what it does:

"Projects a vector on n random hyperplane normals and assigns a binary value to each projection depending on the sign. This divides the data set by each hyperplane and generates a binary hash value in string form, which is being used as a bucket key for storage."

But of course, I could add more of the documentation stuff to the README file. Will do.

pixelogik commented 9 years ago

I added some docs to the README, more to come.