Fix multiple definition of get_column_index.

chrisdubois · chrisdubois · commit dce0d58f666e · 2015-07-28T13:30:15.000-07:00
diff --git a/README.md b/README.md
@@ -17,7 +17,7 @@ To run the following examples you will also need to [register for GraphLab Creat
 
 After that, try running the basic example:
 ```
-python examples/basic.py
+ipython examples/basic.py
 ```
 
 If you want to try a less synthetic example, download the [1TB Criteo dataset](http://labs.criteo.com/downloads/download-terabyte-click-logs/). 
@@ -28,12 +28,12 @@ gzip -cd day_0.gz| head -n 1000000 > criteo-sample.tsv
 
 Next we have a sample script for performing some of the same types of feature engineering that the contest winners have been using:
 ```
-python examples/criteo_process.py
+ipython examples/criteo_process.py
 ```
 
 Train a FFM model on this data.
 ```
-python examples/criteo_sample.py
+ipython examples/criteo_sample.py
 ```
 
 You should see something like the following (which appears to be overfitting in this example):
diff --git a/examples/basic.py b/examples/basic.py
@@ -1,3 +1,4 @@
+import graphlab as gl
 import ffm
 from convert import read_libffm_file
 
@@ -6,7 +7,9 @@
 validfile = 'lib/bigdata.te.txt'
 train = read_libffm_file(trainfile)
 valid = read_libffm_file(validfile)
-valid['features.0'] = None
+
+train['y'] = train['y'].astype(int)
+del train['features.0']
 valid = valid[train.column_names()]
 train.save('examples/small.tr.sframe')
 valid.save('examples/small.te.sframe')
@@ -16,5 +19,5 @@
 # Train a model
 m = ffm.FFM()
 m.fit(train, valid, target='y', features=features, nr_iters=15)
-yhat = m.predict(valid)
-print yhat
+# yhat = m.predict(valid)
+# print yhat
diff --git a/examples/small.te.sframe/dir_archive.ini b/examples/small.te.sframe/dir_archive.ini
@@ -6,4 +6,4 @@ contents=sframe
 [prefixes]
 0000=dir_archive.ini
 0001=objects.bin
-0002=m_e9c35c0d5717c810
+0002=m_2e9354c8c87ae103
diff --git a/examples/small.tr.sframe/dir_archive.ini b/examples/small.tr.sframe/dir_archive.ini
@@ -6,4 +6,4 @@ contents=sframe
 [prefixes]
 0000=dir_archive.ini
 0001=objects.bin
-0002=m_3a9d01aa18c1db1f
+0002=m_9ab3a427b2c5b326
diff --git a/lib/ffm.cpp b/lib/ffm.cpp
@@ -22,11 +22,23 @@
 
 namespace ffm {
 
+size_t get_column_index(graphlab::gl_sframe sf, std::string colname) {
+  const auto colnames = sf.column_names();
+  for (size_t i = 0; i < colnames.size(); ++i) {
+    if (colnames[i] == colname) {
+      return i;
+    }
+  }
+  return -1;
+}
+
+
 namespace {
 
 using namespace std;
 using namespace graphlab;
 
+
 ffm_int const kALIGNByte = 16;
 ffm_int const kALIGN = kALIGNByte/sizeof(ffm_float);
 
@@ -229,11 +241,18 @@ shared_ptr<ffm_model> train(
     }
 
     size_t target_col_idx = get_column_index(tr->sf, tr->target_column); 
+    // logprogress_stream << tr->target_column << " " << get_column_index(tr->sf, tr->target_column) << std::endl;
+    // logprogress_stream << flex_type_enum_to_name(tr->sf.select_column(tr->target_column).dtype()) << std::endl;
+
     std::vector<size_t> feature_col_idxs;
     for (auto col : tr->feature_columns) { 
+      // logprogress_stream << col << " " << get_column_index(tr->sf, col) << std::endl;
+      // logprogress_stream << flex_type_enum_to_name(tr->sf.select_column(col).dtype()) << std::endl;
+
       feature_col_idxs.push_back(get_column_index(tr->sf, col));
     }
 
+    
     for(ffm_int iter = 0; iter < param.nr_iters; iter++)
     {
       ffm_double tr_loss = 0;
@@ -246,8 +265,18 @@ shared_ptr<ffm_model> train(
       for (; it != rsf.end(); ++it, ++i) { 
 
         row_nodes.clear();
-        const std::vector<flexible_type>& row = *it;
+        std::vector<flexible_type> row = *it;
         const auto& yval = row[target_col_idx];
+
+        if (row[target_col_idx].get_type() != flex_type_enum::INTEGER) {
+          log_and_throw("Response must be integer type.");
+        }
+
+        if (row[target_col_idx].get_type() != flex_type_enum::INTEGER) {
+          logprogress_stream << "Column " << target_col_idx << std::endl;
+          logprogress_stream << flex_type_enum_to_name(row[target_col_idx].get_type()) << std::endl;
+          log_and_throw("Response must be integer type.");
+        }
         ffm_float y = (yval.get<flex_int>() > 0) ? 1.0f : -1.0f;
 
         for (size_t col : feature_col_idxs) {
diff --git a/lib/ffm.h b/lib/ffm.h
@@ -12,22 +12,12 @@ namespace ffm
 {
 #endif
 
-
-
 typedef float ffm_float;
 typedef double ffm_double;
 typedef int ffm_int;
 typedef long long ffm_long;
 
-size_t get_column_index(graphlab::gl_sframe sf, std::string colname) {
-  const auto colnames = sf.column_names();
-  for (size_t i = 0; i < colnames.size(); ++i) {
-    if (colnames[i] == colname) {
-      return i;
-    }
-  }
-  return -1;
-}
+size_t get_column_index(graphlab::gl_sframe sf, std::string colname);
 
 typedef graphlab::gl_sarray blah;
 
diff --git a/src/libffm.cpp b/src/libffm.cpp
@@ -19,16 +19,6 @@ using namespace graphlab;
 using namespace std;
 using namespace ffm;
 
-size_t get_column_index(gl_sframe sf, string colname) {
-  const auto colnames = sf.column_names();
-  for (size_t i = 0; i < colnames.size(); ++i) {
-    if (colnames[i] == colname) {
-      return i;
-    }
-  }
-  return -1;
-}
-
 ffm_problem read_sframe(gl_sframe data, std::string target, 
                         std::vector<std::string> features, 
                         size_t max_field_idx, 
