Fix bugs in undocumented sparse matrix conversion functions (#124)

There were bugs in undocumented functions for converting from CSC or CSR formats to
COO format. These bugs did not affect our examples that read sparse matrices from disk
and ran low-rank approximation algorithms; those files have only ever used conversions
from COO to CSC or CSR.

RandBLAS/sparse_data/conversions.hh

@@ -72,7 +72,7 @@ void csc_to_coo(CSCMatrix<T, sint_t1> &csc, COOMatrix<T, sint_t2> &coo) {
     for (int64_t j = 0; j < csc.n_cols; ++j) {
         for (int64_t i = csc.colptr[j]; i < csc.colptr[j+1]; ++i) {
             coo.vals[ell] = csc.vals[ell];
-            coo.rows[ell] = (sint_t2) i;
+            coo.rows[ell] = (sint_t2) csc.rowidxs[i];
             coo.cols[ell] = (sint_t2) j;
             ++ell;
         }
@@ -114,7 +114,7 @@ void csr_to_coo(CSRMatrix<T, sint_t1> &csr, COOMatrix<T, sint_t2> &coo) {
         for (int64_t j = csr.rowptr[i]; j < csr.rowptr[i+1]; ++j) {
             coo.vals[ell] = csr.vals[ell];
             coo.rows[ell] = (sint_t2) i;
-            coo.cols[ell] = (sint_t2) j;
+            coo.cols[ell] = (sint_t2) csr.colidxs[j];
             ++ell;
         }
     }

rtd/source/updates/index.rst

@@ -11,8 +11,6 @@ RandBLAS follows `Semantic Versioning <https://semver.org>`_. Any function docum
 on this website is part of the public API. There are many functions which are not
 part of our public API, but could be added to it if there is user interest. 
 
-RandBLAS is in the 1.0.x release series. The latest version is :ref:`1.0.1 <v10x_patches>`.
-See below for a general overview of each release series.
 
 RandBLAS 1.0
 ------------

test/test_datastructures/test_spmats/test_csc.cc

@@ -34,6 +34,7 @@
 #include <vector>
 
 using namespace RandBLAS::sparse_data;
+using namespace RandBLAS::sparse_data::coo;
 using namespace RandBLAS::sparse_data::csc;
 using namespace test::test_datastructures::test_spmats;
 using namespace RandBLAS::sparse_data::conversions;
@@ -106,7 +107,32 @@ class TestCSC_Conversions : public ::testing::Test {
         delete [] mat_actual;
         return;
     }
+    template <typename T = double>
+    static void test_csc_to_coo_band_diag() {
+        int64_t n = 8;
+        int64_t nnz = 32;
+        std::vector<T> vals{6, -1, -2, -3, -1, 6, -1, -1, -1, 6, -1, -1, -1, -1, 6, -1, -1, 6, -1, -1, -1, -1, 6, -1, -1, -1, 6, -1, -1, -1, -1, 6};
+        std::vector<int64_t> colptr{0, 4, 8, 12, 16, 20, 24, 28, 32};
+        std::vector<int64_t> rowidxs{0, 1, 2, 4, 0, 1, 3, 5, 0, 2, 3, 6, 1, 2, 3, 7, 0, 4, 5, 6, 1, 4, 5, 7, 2, 4, 6, 7, 3, 5, 6, 7};
+        CSCMatrix<T> A_csc(n,n,nnz,vals.data(),rowidxs.data(),colptr.data());
+        COOMatrix<T> A_coo(n,n);
+        csc_to_coo(A_csc, A_coo);
+        std::vector<T> A_dense_coo(n*n);
+        std::vector<T> A_dense_csc(n*n);
+        coo_to_dense(A_coo, Layout::ColMajor, A_dense_coo.data());
+        csc_to_dense(A_csc, Layout::ColMajor, A_dense_csc.data());
+        test::comparison::matrices_approx_equal(
+            Layout::ColMajor, Layout::ColMajor, blas::Op::NoTrans,
+            n, n, A_dense_csc.data(), n, A_dense_coo.data(), n,
+            __PRETTY_FUNCTION__, __FILE__, __LINE__
+        );
+    }
+
 };
+
+TEST_F(TestCSC_Conversions, band) {
+    test_csc_to_coo_band_diag();
+}
 
 TEST_F(TestCSC_Conversions, dense_random_rowmajor) {
     test_csc_from_random_sparsified(Layout::RowMajor, 10, 5, 0.7);

test/test_datastructures/test_spmats/test_csr.cc

@@ -34,6 +34,7 @@
 #include <vector>
 
 using namespace RandBLAS::sparse_data;
+using namespace RandBLAS::sparse_data::coo;
 using namespace RandBLAS::sparse_data::csr;
 using namespace test::test_datastructures::test_spmats;
 using namespace RandBLAS::sparse_data::conversions;
@@ -130,8 +131,34 @@ class TestCSR_Conversions : public ::testing::Test
         delete [] mat_actual;
         return;
     }
+
+    template <typename T = double>
+    static void test_csr_to_coo_band_diagonal() {
+        int64_t n = 8;
+        int64_t nnz = 32;
+        std::vector<T> vals{6, -1, -1, -1, -1, 6, -1, -1, -1, 6, -1, -1, -1, -1, 6, -1, -1, 6, -1, -100, 99, -1, 6, -1, -1, -1, 6, -1, -1, -1, -1, 6};
+        std::vector<int64_t> rowptr{0, 4, 8, 12, 16, 20, 24, 28, 32};
+        std::vector<int64_t> colidxs{0, 1, 2, 4, 0, 1, 3, 5, 0, 2, 3, 6, 1, 2, 3, 7, 0, 4, 5, 6, 1, 4, 5, 7, 2, 4, 6, 7, 3, 5, 6, 7};
+        CSRMatrix<T> A_csr(n,n,nnz,vals.data(),rowptr.data(),colidxs.data());
+        COOMatrix<T> A_coo(n,n);
+        csr_to_coo(A_csr, A_coo);
+        std::vector<T> A_dense_coo(n*n);
+        std::vector<T> A_dense_csr(n*n);
+        coo_to_dense(A_coo, Layout::ColMajor, A_dense_coo.data());
+        csr_to_dense(A_csr, Layout::ColMajor, A_dense_csr.data());
+        test::comparison::matrices_approx_equal(
+            Layout::ColMajor, Layout::ColMajor, blas::Op::NoTrans,
+            n, n, A_dense_csr.data(), n, A_dense_coo.data(), n,
+            __PRETTY_FUNCTION__, __FILE__, __LINE__
+        );
+    }
+
 };
 
+TEST_F(TestCSR_Conversions, band) {
+    test_csr_to_coo_band_diagonal();
+}
+
 TEST_F(TestCSR_Conversions, dense_square_diagonal) {
     test_csr_to_dense_diagonal(3);
  }