Merge branch 'master' into windows_cmake_build

VowpalWabbit · Apr 23, 2019 · 3bdd7bf · 3bdd7bf
2 parents 6c019e7 + bed5b80
commit 3bdd7bf
Show file tree

Hide file tree

Showing 4 changed files with 14 additions and 14 deletions.
diff --git a/.travis.yml b/.travis.yml
@@ -9,4 +9,4 @@ script:
   - docker run -a STDOUT -v `pwd`:/vw -t vowpalwabbit/travis-base:20190202.1 /bin/bash -c "cd /vw && chmod +x ./build-linux.sh && ./build-linux.sh"
 
 after_success:
-  - coveralls --exclude lib --exclude tests --gcov-options '\-lp'
+  - docker run -e TRAVIS_JOB_ID -e TRAVIS_BRANCH -a STDOUT -v `pwd`:/vw -t vowpalwabbit/travis-base:20190202.1 /bin/bash -c "cd /vw && coveralls --include explore --include vowpalwabbit --gcov-options '\-lp'"
diff --git a/README.md b/README.md
@@ -3,7 +3,7 @@
 [![Build Status](https://travis-ci.org/VowpalWabbit/vowpal_wabbit.svg?branch=master)](https://travis-ci.org/VowpalWabbit/vowpal_wabbit)
 [![Windows Build status](https://ci.appveyor.com/api/projects/status/6hqpd9e64h72gybr/branch/master?svg=true)](https://ci.appveyor.com/project/JohnLangford/vowpal-wabbit/branch/master)
 
-[![Coverage Status](https://coveralls.io/repos/JohnLangford/vowpal_wabbit/badge.svg)](https://coveralls.io/r/JohnLangford/vowpal_wabbit)
+[![Coverage Status](https://coveralls.io/repos/github/VowpalWabbit/vowpal_wabbit/badge.svg?branch=master)](https://coveralls.io/github/VowpalWabbit/vowpal_wabbit?branch=master)
 [![Total Alerts](https://img.shields.io/lgtm/alerts/g/JohnLangford/vowpal_wabbit.svg?logo=lgtm&logoWidth=18)](https://lgtm.com/projects/g/JohnLangford/vowpal_wabbit/alerts/)
 [![Gitter chat](https://badges.gitter.im/VowpalWabbit.png)](https://gitter.im/VowpalWabbit)
 

diff --git a/test/train-sets/ref/rcv1_raw_cb_explore.stderr b/test/train-sets/ref/rcv1_raw_cb_explore.stderr
@@ -10,20 +10,20 @@ Reading datafile = train-sets/rcv1_raw_cb_small.vw
 num sources = 1
 average  since         example        example  current  current  current
 loss     last          counter         weight    label  predict features
-0.050000 0.050000            1            1.0        1 1:0.975000      280
-0.025000 0.000000            2            2.0        1 2:0.975000      598
-0.012500 0.000000            4            4.0        1 2:0.975000      784
-0.250000 0.487500            8            8.0        1 2:0.975000      850
-0.368750 0.487500           16           16.0        1 2:0.975000      118
-0.245312 0.121875           32           32.0        1 1:0.975000      166
-0.216406 0.187500           64           64.0        2 1:0.975000      340
-0.216406 0.216406          128          128.0        1 2:0.975000      610
-0.300195 0.383984          256          256.0        2 2:0.975000      712
-0.333984 0.367773          512          512.0        2 1:0.975000      424
+1.950000 1.950000            1            1.0        1 1:0.975000      280
+0.975000 0.000000            2            2.0        1 2:0.975000      598
+0.487500 0.000000            4            4.0        1 2:0.975000      784
+0.250000 0.012500            8            8.0        1 2:0.975000      850
+0.618750 0.987500           16           16.0        1 2:0.975000      118
+0.318750 0.018750           32           32.0        1 1:0.975000      166
+0.321875 0.325000           64           64.0        2 1:0.975000      340
+0.244922 0.167969          128          128.0        1 2:0.975000      610
+0.231250 0.217578          256          256.0        2 2:0.975000      712
+0.216992 0.202734          512          512.0        2 1:0.975000      424
 
 finished run
 number of examples = 1000
 weighted example sum = 1000.000000
 weighted label sum = 0.000000
-average loss = 0.362700
+average loss = 0.160650
 total feature number = 440020
diff --git a/vowpalwabbit/cb_explore.cc b/vowpalwabbit/cb_explore.cc
@@ -241,7 +241,7 @@ void output_example(vw& all, cb_explore& data, example& ec, CB::label& ld)
 
   if ((c.known_cost = get_observed_cost(ld)) != nullptr)
     for (uint32_t i = 0; i < ec.pred.a_s.size(); i++)
-      loss += get_unbiased_cost(c.known_cost, c.pred_scores, i) * ec.pred.a_s[i].score;
+      loss += get_unbiased_cost(c.known_cost, c.pred_scores, i+1) * ec.pred.a_s[i].score;
 
   all.sd->update(ec.test_only, get_observed_cost(ld) != nullptr, loss, 1.f, ec.num_features);