sc::SCMatrix(3)

💖 sc::SCMatrix(3) — the MPQC manual page, bestie 💖

okay so this is sc::SCMatrix(3) from the MPQC manual, and it’s giving abstract-matrix main character energy ✨

🌸 NAME 🌸

sc::SCMatrix — the abstract base class for general double valued n by m matrices. That’s the whole name vibe, and yes, it’s iconic 💅

✨ SYNOPSIS ✨

#include <abstract.h>

Inherits sc::DescribedClass.

Inherited by sc::BlockedSCMatrix, sc::DistSCMatrix, sc::LocalSCMatrix,
and sc::ReplSCMatrix.

🎀 Public Types 🎀

• enum Transform { NormalTransform = 0, TransposeTransform = 1 } — the transform enum, with NormalTransform = 0 and TransposeTransform = 1, exactly as the page says 🦋

💅 Public Member Functions 💅

• SCMatrix (const RefSCDimension &, const RefSCDimension &, SCMatrixKit *) — the constructor, taking two RefSCDimensions and an SCMatrixKit *.
• virtual void save (StateOut &) — save and restore this in an implementation independent way 💾
• virtual void restore (StateIn &) — the restore half of that implementation-independent save/restore situation.
• Ref< SCMatrixKit > kit () const — return the SCMatrixKit used to create this object.
• int nrow () const — return the number of rows.
• int ncol () const — return the number of columns.
• virtual double maxabs () const — return the maximum absolute value element.
• virtual void randomize () — assign each element to a random number between -1 and 1.
• void assign (double val) — set all elements to val.
• void assign (const double *m) — assign element i, j to m[ir*nrow()+j] — yes, exactly like that.
• void assign (const double **m) — assign element i, j to m[i][j].
• void assign (SCMatrix *m) — make this have the same elements as m. The dimensions must match.
• virtual void assign_val (double val) — overridden to implement the assign members.
• virtual void assign_p (const double *m) — the page gives no extra prose here.
• virtual void assign_pp (const double **m) — the page gives no extra prose here either.
• virtual void assign_r (SCMatrix *m) — same assign family, still no extra prose in the page.
• virtual void convert (double *) const — like the assign members, but these write values to the arguments.
• virtual void convert (double **) const — the double-pointer convert sibling; no extra prose in the page.
• virtual void convert (SCMatrix *) — convert an SCMatrix of a different specialization to this specialization and possibly accumulate the data.
• virtual void convert_accumulate (SCMatrix *) — the page gives no extra prose here.
• virtual void scale (double val) — multiply all elements by val.
• virtual void scale_diagonal (double val) — scale the diagonal elements by val.
• virtual void shift_diagonal (double val) — shift the diagonal elements by val.
• virtual void unit () — make this equal to the unit matrix.
• virtual SCMatrix * copy () — return a matrix with the same dimension and same elements.
• virtual SCMatrix * clone () — return a matrix with the same dimension but uninitialized memory.
• RefSCDimension rowdim () const — return the row or column dimension; this is the row half of that pair.
• RefSCDimension coldim () const — return the row or column dimension; this is the column half of that pair.
• virtual double get_element (int, int) const =0 — return or modify an element. Pure virtual.
• virtual void set_element (int, int, double)=0 — the set half of the “return or modify an element” note. Pure virtual.
• virtual void accumulate_element (int, int, double)=0 — the accumulate half of that element note. Pure virtual.
• virtual SCMatrix * get_subblock (int br, int er, int bc, int ec)=0 — return a subblock of this. Pure virtual, with the full definition below.
• virtual void assign_subblock (SCMatrix *m, int, int, int, int, int=0, int=0)=0 — assign m to a subblock of this. Pure virtual.
• virtual void accumulate_subblock (SCMatrix *m, int, int, int, int, int=0, int=0)=0 — sum m into a subblock of this. Pure virtual.
• virtual SCVector * get_row (int i)=0 — return a row or column of this. Pure virtual.
• virtual SCVector * get_column (int i)=0 — return a row or column of this. Pure virtual.
• virtual void assign_row (SCVector *v, int i)=0 — assign v to a row or column of this. Pure virtual.
• virtual void assign_column (SCVector *v, int i)=0 — assign v to a row or column of this. Pure virtual.
• virtual void accumulate_row (SCVector *v, int i)=0 — sum v to a row or column of this. Pure virtual.
• virtual void accumulate_column (SCVector *v, int i)=0 — sum v to a row or column of this. Pure virtual.
• virtual void accumulate (const SCMatrix *m)=0 — sum m into this. Pure virtual.
• virtual void accumulate (const SymmSCMatrix *m)=0 — the SymmSCMatrix accumulate overload. Pure virtual.
• virtual void accumulate (const DiagSCMatrix *m)=0 — the DiagSCMatrix accumulate overload. Pure virtual.
• virtual void accumulate (const SCVector *)=0 — the SCVector accumulate overload. Pure virtual.
• virtual void accumulate_outer_product (SCVector *, SCVector *)=0 — sum into this the products of various vectors or matrices. Pure virtual.
• void accumulate_product (SCMatrix *m1, SCMatrix *m2) — product-accumulate overload; the page gives no extra prose for it.
• void accumulate_product (SCMatrix *m1, SymmSCMatrix *m2) — another product-accumulate overload.
• void accumulate_product (SCMatrix *m1, DiagSCMatrix *m2) — another product-accumulate overload.
• void accumulate_product (SymmSCMatrix *m1, SCMatrix *m2) — another product-accumulate overload.
• void accumulate_product (DiagSCMatrix *m1, SCMatrix *m2) — another product-accumulate overload.
• void accumulate_product (SymmSCMatrix *m1, SymmSCMatrix *m2) — another product-accumulate overload.
• virtual void accumulate_product_rr (SCMatrix *, SCMatrix *)=0 — pure virtual product-accumulate overload.
• virtual void accumulate_product_rs (SCMatrix *, SymmSCMatrix *) — product-accumulate overload; no extra prose in the page.
• virtual void accumulate_product_rd (SCMatrix *, DiagSCMatrix *) — product-accumulate overload; no extra prose in the page.
• virtual void accumulate_product_sr (SymmSCMatrix *, SCMatrix *) — product-accumulate overload; no extra prose in the page.
• virtual void accumulate_product_dr (DiagSCMatrix *, SCMatrix *) — product-accumulate overload; no extra prose in the page.
• virtual void accumulate_product_ss (SymmSCMatrix *, SymmSCMatrix *) — product-accumulate overload; no extra prose in the page.
• virtual void transpose_this ()=0 — transpose this. Pure virtual.
• virtual double trace ()=0 — return the trace. Pure virtual.
• virtual double invert_this ()=0 — invert this. Pure virtual.
• virtual double determ_this ()=0 — return the determinant of this. this is overwritten. Pure virtual.
• virtual void svd_this (SCMatrix *U, DiagSCMatrix *sigma, SCMatrix *V) — compute the singular value decomposition for this, possibly destroying this.
• virtual double solve_this (SCVector *)=0 — pure virtual; the page gives no extra prose here.
• virtual void gen_invert_this () — virtual; the page gives no extra prose here.
• virtual void schmidt_orthog (SymmSCMatrix *, int n)=0 — Schmidt orthogonalize this. Pure virtual, with the full definition below.
• virtual int schmidt_orthog_tol (SymmSCMatrix *, double tol, double *res=0)=0 — Schmidt orthogonalize this. Pure virtual, with the full definition below.
• virtual void element_op (const Ref< SCElementOp > &)=0 — perform the element operation op on each element of this. Pure virtual.
• virtual void element_op (const Ref< SCElementOp2 > &, SCMatrix *)=0 — the two-matrix element operation overload. Pure virtual.
• virtual void element_op (const Ref< SCElementOp3 > &, SCMatrix *, SCMatrix *)=0 — the three-matrix element operation overload. Pure virtual.
• void print (std::ostream &o=ExEnv::out0()) const — print out the matrix.
• void print (const char *title=0, std::ostream &out=ExEnv::out0(), int=10) const — print out the matrix, with the title and stream and int defaults exactly as shown.
• virtual void vprint (const char *title=0, std::ostream &out=ExEnv::out0(), int=10) const =0 — pure virtual print member; the page gives no extra prose here.
• Ref< MessageGrp > messagegrp () const — returns the message group used by the matrix kit.
• virtual Ref< SCMatrixSubblockIter > local_blocks (SCMatrixSubblockIter::Access)=0 — returns iterators for the local (rapidly accessible) blocks used in this matrix. Pure virtual.
• virtual Ref< SCMatrixSubblockIter > all_blocks (SCMatrixSubblockIter::Access)=0 — returns iterators for the all blocks used in this matrix. Pure virtual.

🔒 Protected Attributes 🔒

• RefSCDimension d1 — protected attribute, exactly as listed.
• RefSCDimension d2 — protected attribute, exactly as listed.
• Ref< SCMatrixKit > kit_ — protected attribute, exactly as listed.

🌈 Additional Inherited Members 🌈

The manual puts this heading here and then moves on to the Detailed Description. No extra members are spelled out under it, no cap.

🌸 Detailed Description 🌸

The SCMatrix class is the abstract base class for general double valued n by m matrices. That’s the whole mood board 💖

For symmetric matrices use SymmSCMatrix, and for diagonal matrices use DiagSCMatrix. Very important, bestie, don’t mix up your matrix specializations ✨

🎀 Member Function Documentation 🎀

💫 virtual SCMatrix* sc::SCMatrix::get_subblock (int br, int er, int bc, int ec) [pure virtual] 💫

Return a subblock of this. The subblock is defined as the rows starting at br and ending at er, and the columns beginning at bc and ending at ec. So yes: rows br through er, columns bc through ec 🫶

Implemented in sc::BlockedSCMatrix, sc::ReplSCMatrix, sc::DistSCMatrix, and sc::LocalSCMatrix.

💫 virtual void sc::SCMatrix::schmidt_orthog (SymmSCMatrix *, int n) [pure virtual] 💫

Schmidt orthogonalize this. S is the overlap matrix. n is the number of columns to orthogonalize. That’s the whole note, and it’s lowkey important 💅

Implemented in sc::BlockedSCMatrix, sc::ReplSCMatrix, sc::DistSCMatrix, and sc::LocalSCMatrix.

💫 virtual int sc::SCMatrix::schmidt_orthog_tol (SymmSCMatrix *, double tol, double * res = 0) [pure virtual] 💫

Schmidt orthogonalize this. S is the overlap matrix. tol is the tolerance. The number of linearly independent vectors is returned. The return value is literally the count of linearly independent vectors, slay 🦋

Implemented in sc::BlockedSCMatrix, sc::ReplSCMatrix, sc::DistSCMatrix, and sc::LocalSCMatrix.

👑 Author 👑

Generated automatically by Doxygen for MPQC from the source code. The robots did this one, bestie 🤖✨

---

This page is from MPQC version 2.3.1, dated Sun Oct 4 2020, and it’s the sc::SCMatrix(3) manual page. Stay abstract, stay iconic 💜🫶