Skip to content

CF toolkit: Use rankUpdate() to calculate XtX in ALS and WALS #24

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 5 commits into
base: master
Choose a base branch
from
Open

CF toolkit: Use rankUpdate() to calculate XtX in ALS and WALS #24

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
16967fe
Fixed typo in regnormal code
Apr 29, 2014
8e49446
Merge branch 'master' of https://github.com/GraphChi/graphchi-cpp
May 5, 2014
162a6ed
LibFM: fixed typo in load last_item_bias code
May 5, 2014
5c403ad
Merge branch 'master' of https://github.com/GraphChi/graphchi-cpp
May 14, 2014
07e530e
ALS/WALS: use rankUpdate() to update XtX
May 14, 2014
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion toolkits/collaborative_filtering/als.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -131,7 +131,7 @@ struct ALSVerticesInMemProgram : public GraphChiProgram<VertexDataType, EdgeData
float observation = vertex.edge(e)->get_data();
vertex_data & nbr_latent = latent_factors_inmem[vertex.edge(e)->vertex_id()];
Xty += nbr_latent.pvec * observation;
XtX.triangularView<Eigen::Upper>() += nbr_latent.pvec * nbr_latent.pvec.transpose();
XtX.selfadjointView<Eigen::Upper>().rankUpdate(nbr_latent.pvec);
if (compute_rmse) {
double prediction;
rmse_vec[omp_get_thread_num()] += als_predict(vdata, nbr_latent, observation, prediction);
Expand Down
2 changes: 1 addition & 1 deletion toolkits/collaborative_filtering/libfm.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -314,7 +314,7 @@ int main(int argc, const char ** argv) {
vec user_bias = load_matrix_market_vector(training +"_U_bias.mm", false, true);
vec item_bias = load_matrix_market_vector(training +"_V_bias.mm", false, true);
vec time_bias = load_matrix_market_vector(training+ "_T_bias.mm", false, true);
vec last_item_bias = load_matrix_market_vector(training+"_L_bias.m", false, true);
vec last_item_bias = load_matrix_market_vector(training+"_L_bias.mm", false, true);
for (uint i=0; i<M+N+K+M; i++){
if (i < M)
latent_factors_inmem[i].bias = user_bias[i];
Expand Down
2 changes: 1 addition & 1 deletion toolkits/collaborative_filtering/wals.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -119,7 +119,7 @@ struct WALSVerticesInMemProgram : public GraphChiProgram<VertexDataType, EdgeDat
const edge_data & edge = vertex.edge(e)->get_data();
vertex_data & nbr_latent = latent_factors_inmem[vertex.edge(e)->vertex_id()];
Xty += nbr_latent.pvec * edge.weight * edge.time;
XtX.triangularView<Eigen::Upper>() += nbr_latent.pvec * nbr_latent.pvec.transpose() * edge.time;
XtX.selfadjointView<Eigen::Upper>().rankUpdate(nbr_latent.pvec, edge.time);
if (compute_rmse) {
double prediction;
rmse_vec[omp_get_thread_num()] += wals_predict(vdata, nbr_latent, edge.weight, prediction) * edge.time;
Expand Down