Hands-On Deep Learning with Go

Preparing our data – GroupLens movie ratings

We are using the GroupLens dataset. It contains a collection of users, movies, and ratings collected from MovieLens (http://www.movielens.org), and is run by a number of academic researchers at the University of Minnesota.

We need to parse the ratings.dat file, delimited with a colon, for userids, ratings, and movieids. We can then match up movieids with those in movies.dat.

First, let's look at the code we need to build our index of movies:

package mainimport (  // "github.com/aotimme/rbm"  "fmt"  "log"  "math"  "strconv"  "github.com/yunabe/easycsv"  g "gorgonia.org/gorgonia"  "gorgonia.org/tensor")var datasetfilename string = "dataset/cleanratings.csv"var movieindexfilename string = "dataset/cleanmovies.csv" ...