Eigen's documentation for sparse matrices states that "the elements of a given inner vector are guaranteed to be always sorted by increasing inner indices" (https://eigen.tuxfamily.org/dox/group__TutorialSparse.html). This fact allows more efficient access and insert operations, but it causes issues when I try to use an Eigen::Map for sparse matrix types when the underlying storage does not follow the above ordering rules (https://eigen.tuxfamily.org/dox/classEigen_1_1Map_3_01SparseMatrixType_01_4.html). However, I realized that not all member functions of Eigen::SparseMatrix rely on the assumption of increasing inner indices; for example, the toDense method. Is there a comprehensive list of such functions? If they are not too many, I would then try to reimplement them for my codebase.

As a side note, I would be happy if there was an option to enable different inner index orders (increasing/decreasing/unordered) for sparse matrices