Fix r interaction constraints (#5543)

* Unify the parsing code.

* Cleanup.

R-package/tests/testthat/test_interaction_constraints.R

@@ -34,5 +34,22 @@ test_that("interaction constraints for regression", {
   expect_true({
     test1 & test2
   }, "Interaction Contraint Satisfied")
-  
+})
+
+test_that("interaction constraints scientific representation", {
+  rows <- 10
+  ## When number exceeds 1e5, R paste function uses scientific representation.
+  ## See: https://github.com/dmlc/xgboost/issues/5179
+  cols <- 1e5+10
+
+  d <- matrix(rexp(rows, rate=.1), nrow=rows, ncol=cols)
+  y <- rnorm(rows)
+
+  dtrain <- xgb.DMatrix(data=d, info = list(label=y))
+  inc <- list(c(seq.int(from = 0, to = cols, by = 1)))
+
+  with_inc <- xgb.train(data=dtrain, tree_method='hist',
+                        interaction_constraints=inc, nrounds=10)
+  without_inc <- xgb.train(data=dtrain, tree_method='hist', nrounds=10)
+  expect_equal(xgb.save.raw(with_inc), xgb.save.raw(without_inc))
 })

src/tree/constraints.cc

@@ -6,6 +6,7 @@
 #include <vector>
 
 #include "xgboost/span.h"
+#include "xgboost/json.h"
 #include "constraints.h"
 #include "param.h"
 
@@ -27,14 +28,11 @@ void FeatureInteractionConstraintHost::Reset() {
   if (!enabled_) {
     return;
   }
-  // Parse interaction constraints
-  std::istringstream iss(this->interaction_constraint_str_);
-  dmlc::JSONReader reader(&iss);
-  // Read std::vector<std::vector<bst_uint>> first and then
-  //   convert to std::vector<std::unordered_set<bst_uint>>
-  std::vector<std::vector<bst_uint>> tmp;
+  // Read std::vector<std::vector<bst_feature_t>> first and then
+  //   convert to std::vector<std::unordered_set<bst_feature_t>>
+  std::vector<std::vector<bst_feature_t>> tmp;
   try {
-    reader.Read(&tmp);
+    ParseInteractionConstraint(this->interaction_constraint_str_, &tmp);
   } catch (dmlc::Error const &e) {
     LOG(FATAL) << "Failed to parse feature interaction constraint:\n"
                << this->interaction_constraint_str_ << "\n"

src/tree/constraints.cu

@@ -7,9 +7,7 @@
 #include <thrust/iterator/counting_iterator.h>
 
 #include <algorithm>
-#include <bitset>
 #include <string>
-#include <sstream>
 #include <set>
 
 #include "xgboost/logging.h"
@@ -18,14 +16,13 @@
 #include "param.h"
 #include "../common/device_helpers.cuh"
 
-
 namespace xgboost {
 
-size_t FeatureInteractionConstraint::Features() const {
+size_t FeatureInteractionConstraintDevice::Features() const {
   return d_sets_ptr_.size() - 1;
 }
 
-void FeatureInteractionConstraint::Configure(
+void FeatureInteractionConstraintDevice::Configure(
     tree::TrainParam const& param, int32_t const n_features) {
   has_constraint_ = true;
   if (param.interaction_constraints.length() == 0) {
@@ -33,13 +30,11 @@ void FeatureInteractionConstraint::Configure(
     return;
   }
   // --- Parse interaction constraints
-  std::istringstream iss(param.interaction_constraints);
-  dmlc::JSONReader reader(&iss);
   // Interaction constraints parsed from string parameter.  After
   // parsing, this looks like {{0, 1, 2}, {2, 3 ,4}}.
-  std::vector<std::vector<int32_t>> h_feature_constraints;
+  std::vector<std::vector<bst_feature_t>> h_feature_constraints;
   try {
-    reader.Read(&h_feature_constraints);
+    ParseInteractionConstraint(param.interaction_constraints, &h_feature_constraints);
   } catch (dmlc::Error const& e) {
     LOG(FATAL) << "Failed to parse feature interaction constraint:\n"
                << param.interaction_constraints << "\n"
@@ -68,13 +63,13 @@ void FeatureInteractionConstraint::Configure(
 
   // Represent constraints as CSR format, flatten is the value vector,
   // ptr is row_ptr vector in CSR.
-  std::vector<int32_t> h_feature_constraints_flatten;
+  std::vector<uint32_t> h_feature_constraints_flatten;
   for (auto const& constraints : h_feature_constraints) {
-    for (int32_t c : constraints) {
+    for (uint32_t c : constraints) {
       h_feature_constraints_flatten.emplace_back(c);
     }
   }
-  std::vector<int32_t> h_feature_constraints_ptr;
+  std::vector<size_t> h_feature_constraints_ptr;
   size_t n_features_in_constraints = 0;
   h_feature_constraints_ptr.emplace_back(n_features_in_constraints);
   for (auto const& v : h_feature_constraints) {
@@ -130,13 +125,13 @@ void FeatureInteractionConstraint::Configure(
   s_result_buffer_ = dh::ToSpan(result_buffer_);
 }
 
-FeatureInteractionConstraint::FeatureInteractionConstraint(
+FeatureInteractionConstraintDevice::FeatureInteractionConstraintDevice(
     tree::TrainParam const& param, int32_t const n_features) :
     has_constraint_{true}, n_sets_{0} {
   this->Configure(param, n_features);
 }
 
-void FeatureInteractionConstraint::Reset() {
+void FeatureInteractionConstraintDevice::Reset() {
   for (auto& node : node_constraints_storage_) {
     thrust::fill(node.begin(), node.end(), 0);
   }
@@ -153,7 +148,7 @@ __global__ void ClearBuffersKernel(
   }
 }
 
-void FeatureInteractionConstraint::ClearBuffers() {
+void FeatureInteractionConstraintDevice::ClearBuffers() {
   CHECK_EQ(output_buffer_bits_.Size(), input_buffer_bits_.Size());
   CHECK_LE(feature_buffer_.Size(), output_buffer_bits_.Size());
   uint32_t constexpr kBlockThreads = 256;
@@ -164,7 +159,7 @@ void FeatureInteractionConstraint::ClearBuffers() {
       output_buffer_bits_, input_buffer_bits_);
 }
 
-common::Span<bst_feature_t> FeatureInteractionConstraint::QueryNode(int32_t node_id) {
+common::Span<bst_feature_t> FeatureInteractionConstraintDevice::QueryNode(int32_t node_id) {
   if (!has_constraint_) { return {}; }
   CHECK_LT(node_id, s_node_constraints_.size());
 
@@ -203,7 +198,7 @@ __global__ void QueryFeatureListKernel(LBitField64 node_constraints,
   result_buffer_output &= result_buffer_input;
 }
 
-common::Span<bst_feature_t> FeatureInteractionConstraint::Query(
+common::Span<bst_feature_t> FeatureInteractionConstraintDevice::Query(
     common::Span<bst_feature_t> feature_list, int32_t nid) {
   if (!has_constraint_ || nid == 0) {
     return feature_list;
@@ -250,8 +245,8 @@ __global__ void RestoreFeatureListFromSetsKernel(
     LBitField64 feature_buffer,
 
     bst_feature_t fid,
-    common::Span<int32_t> feature_interactions,
-    common::Span<int32_t> feature_interactions_ptr,  // of size n interaction set + 1
+    common::Span<bst_feature_t> feature_interactions,
+    common::Span<size_t> feature_interactions_ptr,  // of size n interaction set + 1
 
     common::Span<bst_feature_t> interactions_list,
     common::Span<size_t> interactions_list_ptr) {
@@ -302,7 +297,7 @@ __global__ void InteractionConstraintSplitKernel(LBitField64 feature,
   }
 }
 
-void FeatureInteractionConstraint::Split(
+void FeatureInteractionConstraintDevice::Split(
     bst_node_t node_id, bst_feature_t feature_id, bst_node_t left_id, bst_node_t right_id) {
   if (!has_constraint_) { return; }
   CHECK_NE(node_id, left_id)

src/tree/constraints.cuh

@@ -88,18 +88,18 @@ struct ValueConstraint {
 };
 
 // Feature interaction constraints built for GPU Hist updater.
-struct FeatureInteractionConstraint {
+struct FeatureInteractionConstraintDevice {
  protected:
   // Whether interaction constraint is used.
   bool has_constraint_;
   // n interaction sets.
-  int32_t n_sets_;
+  size_t n_sets_;
 
   // The parsed feature interaction constraints as CSR.
-  dh::device_vector<int32_t> d_fconstraints_;
-  common::Span<int32_t> s_fconstraints_;
-  dh::device_vector<int32_t> d_fconstraints_ptr_;
-  common::Span<int32_t> s_fconstraints_ptr_;
+  dh::device_vector<bst_feature_t> d_fconstraints_;
+  common::Span<bst_feature_t> s_fconstraints_;
+  dh::device_vector<size_t> d_fconstraints_ptr_;
+  common::Span<size_t> s_fconstraints_ptr_;
   /* Interaction sets for each feature as CSR.  For an input like:
    * [[0, 1], [1, 2]], this will have values:
    *
@@ -141,11 +141,11 @@ struct FeatureInteractionConstraint {
 
  public:
   size_t Features() const;
-  FeatureInteractionConstraint() = default;
+  FeatureInteractionConstraintDevice() = default;
   void Configure(tree::TrainParam const& param, int32_t const n_features);
-  FeatureInteractionConstraint(tree::TrainParam const& param, int32_t const n_features);
-  FeatureInteractionConstraint(FeatureInteractionConstraint const& that) = default;
-  FeatureInteractionConstraint(FeatureInteractionConstraint&& that) = default;
+  FeatureInteractionConstraintDevice(tree::TrainParam const& param, int32_t const n_features);
+  FeatureInteractionConstraintDevice(FeatureInteractionConstraintDevice const& that) = default;
+  FeatureInteractionConstraintDevice(FeatureInteractionConstraintDevice&& that) = default;
   /*! \brief Reset before constructing a new tree. */
   void Reset();
   /*! \brief Return a list of features given node id */

src/tree/param.cc

@@ -5,6 +5,7 @@
 #include <vector>
 #include <utility>
 
+#include "xgboost/json.h"
 #include "param.h"
 
 namespace std {
@@ -79,3 +80,31 @@ std::istream &operator>>(std::istream &is, std::vector<int> &t) {
   return is;
 }
 }  // namespace std
+
+namespace xgboost {
+void ParseInteractionConstraint(
+    std::string const &constraint_str,
+    std::vector<std::vector<bst_feature_t>> *p_out) {
+  auto &out = *p_out;
+  auto j_inc = Json::Load({constraint_str.c_str(), constraint_str.size()});
+  auto const &all = get<Array>(j_inc);
+  out.resize(all.size());
+  for (size_t i = 0; i < all.size(); ++i) {
+    auto const &set = get<Array const>(all[i]);
+    for (auto const &v : set) {
+      if (XGBOOST_EXPECT(IsA<Integer>(v), true)) {
+        uint32_t u = static_cast<uint32_t const>(get<Integer const>(v));
+        out[i].emplace_back(u);
+      } else if (IsA<Number>(v)) {
+        double d = get<Number const>(v);
+        CHECK_EQ(std::floor(d), d)
+            << "Found floating point number in interaction constraints";
+        out[i].emplace_back(static_cast<uint32_t const>(d));
+      } else {
+        LOG(FATAL) << "Unknown value type for interaction constraint:"
+                   << v.GetValue().TypeStr();
+      }
+    }
+  }
+}
+}  // namespace xgboost

src/tree/param.h

@@ -483,8 +483,21 @@ struct SplitEntryContainer {
 };
 
 using SplitEntry = SplitEntryContainer<GradStats>;
-
 }  // namespace tree
+
+/*
+ * \brief Parse the interaction constraints from string.
+ * \param constraint_str String storing the interfaction constraints:
+ *
+ *  Example input string:
+ *
+ *    "[[1, 2], [3, 4]]""
+ *
+ * \param p_out Pointer to output
+ */
+void ParseInteractionConstraint(
+    std::string const &constraint_str,
+    std::vector<std::vector<xgboost::bst_feature_t>> *p_out);
 }  // namespace xgboost
 
 // define string serializer for vector, to get the arguments

src/tree/updater_gpu_hist.cu

@@ -436,7 +436,7 @@ struct GPUHistMakerDevice {
   common::Monitor monitor;
   std::vector<ValueConstraint> node_value_constraints;
   common::ColumnSampler column_sampler;
-  FeatureInteractionConstraint interaction_constraints;
+  FeatureInteractionConstraintDevice interaction_constraints;
 
   using ExpandQueue =
       std::priority_queue<ExpandEntry, std::vector<ExpandEntry>,

tests/cpp/tree/test_constraints.cu

@@ -15,12 +15,12 @@
 namespace xgboost {
 namespace {
 
-struct FConstraintWrapper : public FeatureInteractionConstraint {
+struct FConstraintWrapper : public FeatureInteractionConstraintDevice {
   common::Span<LBitField64> GetNodeConstraints() {
-    return FeatureInteractionConstraint::s_node_constraints_;
+    return FeatureInteractionConstraintDevice::s_node_constraints_;
   }
   FConstraintWrapper(tree::TrainParam param, bst_feature_t n_features) :
-      FeatureInteractionConstraint(param, n_features) {}
+      FeatureInteractionConstraintDevice(param, n_features) {}
 
   dh::device_vector<bst_feature_t> const& GetDSets() const {
     return d_sets_;