Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

still need to test row merge

Browse Source
This commit is contained in:
tqchen 2014-11-19 11:44:24 -08:00
parent da54f5e5d8
commit 55e62a7120
4 changed files with 6 additions and 23 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
multi-node/col-split/README.md
View File

@ -1,14 +1,14 @@
Column Split Version of XGBoost Distributed XGBoost: Column Split Version
==== ====
* run ```bash run-mushroom.sh``` * run ```bash run-mushroom.sh```
Steps to use column split version How to Use
==== ====
* First split the data by column, * First split the data by column,
* In the config, specify data file as containing a wildcard %d, where %d is the rank of the node, each node will load their part of data * In the config, specify data file as containing a wildcard %d, where %d is the rank of the node, each node will load their part of data
* Enable column split mode by ```dsplit=col``` * Enable column split mode by ```dsplit=col```
Note on the Column Split Version Notes
==== ====
* The code is multi-threaded, so you want to run one xgboost-mpi per node * The code is multi-threaded, so you want to run one xgboost-mpi per node
* The code will work correctly as long as union of each column subset is all the columns we are interested in. * The code will work correctly as long as union of each column subset is all the columns we are interested in.

0
multi-node/col-split/runexp-mpi.sh → multi-node/col-split/mushroom-col.sh Normal file → Executable file
View File

19
multi-node/col-split/run-mushroom.sh
View File

@ -1,19 +0,0 @@
#!/bin/bash
if [[ $# -ne 1 ]]
then
echo "Usage: nprocess"
exit -1
fi
rm -rf train.col*
k=$1
# split the lib svm file into k subfiles
python splitsvm.py ../../demo/data/agaricus.txt.train train $k
# run xgboost mpi
mpirun -n $k ../../xgboost-mpi mushroom-col.conf updater=distcol silent=0
# the model can be directly loaded by single machine xgboost solver, as usuall
../../xgboost mushroom-col.conf task=dump model_in=0002.model fmap=../../demo/data/featmap.txt name_dump=dump.nice.$k.txt
cat dump.nice.$k.txt

4
src/learner/learner-inl.hpp
View File

@ -92,7 +92,7 @@ class BoostLearner {
if (!strcmp(name, "silent")) silent = atoi(val); if (!strcmp(name, "silent")) silent = atoi(val);
if (!strcmp(name, "dsplit")) { if (!strcmp(name, "dsplit")) {
if (!strcmp(val, "col")) { if (!strcmp(val, "col")) {
this->SetParam("updater", "distcol,prune"); this->SetParam("updater", "distcol");
distributed_mode = 1; distributed_mode = 1;
} else if (!strcmp(val, "row")) { } else if (!strcmp(val, "row")) {
this->SetParam("updater", "grow_histmaker,prune"); this->SetParam("updater", "grow_histmaker,prune");
@ -104,6 +104,8 @@ class BoostLearner {
if (!strcmp(name, "part_load_col")) part_load_col = atoi(val); if (!strcmp(name, "part_load_col")) part_load_col = atoi(val);
if (!strcmp(name, "prob_buffer_row")) { if (!strcmp(name, "prob_buffer_row")) {
prob_buffer_row = static_cast<float>(atof(val)); prob_buffer_row = static_cast<float>(atof(val));
utils::Check(distributed_mode == 0,
"prob_buffer_row can only be used in single node mode so far");
this->SetParam("updater", "grow_colmaker,refresh,prune"); this->SetParam("updater", "grow_colmaker,refresh,prune");
} }
if (!strcmp(name, "eval_metric")) evaluator_.AddEval(val); if (!strcmp(name, "eval_metric")) evaluator_.AddEval(val);