1 // A simple quickref for Eigen. Add anything that's missing.
2 // Main author: Keir Mierle
6 Matrix<double, 3, 3> A; // Fixed rows and cols. Same as Matrix3d.
7 Matrix<double, 3, Dynamic> B; // Fixed rows, dynamic cols.
8 Matrix<double, Dynamic, Dynamic> C; // Full dynamic. Same as MatrixXd.
9 Matrix<double, 3, 3, RowMajor> E; // Row major; default is column-major.
10 Matrix3f P, Q, R; // 3x3 float matrix.
11 Vector3f x, y, z; // 3x1 float matrix.
12 RowVector3f a, b, c; // 1x3 float matrix.
13 VectorXd v; // Dynamic column vector of doubles
17 // Eigen // Matlab // comments
18 x.size() // length(x) // vector size
19 C.rows() // size(C,1) // number of rows
20 C.cols() // size(C,2) // number of columns
21 x(i) // x(i+1) // Matlab is 1-based
22 C(i,j) // C(i+1,j+1) //
24 A.resize(4, 4); // Runtime error if assertions are on.
25 B.resize(4, 9); // Runtime error if assertions are on.
26 A.resize(3, 3); // Ok; size didn't change.
27 B.resize(3, 9); // Ok; only dynamic cols changed.
29 A << 1, 2, 3, // Initialize A. The elements can also be
30 4, 5, 6, // matrices, which are stacked along cols
31 7, 8, 9; // and then the rows are stacked.
32 B << A, A, A; // B is three horizontally stacked A's.
33 A.fill(10); // Fill A with all 10's.
36 MatrixXd::Identity(rows,cols) // eye(rows,cols)
37 C.setIdentity(rows,cols) // C = eye(rows,cols)
38 MatrixXd::Zero(rows,cols) // zeros(rows,cols)
39 C.setZero(rows,cols) // C = zeros(rows,cols)
40 MatrixXd::Ones(rows,cols) // ones(rows,cols)
41 C.setOnes(rows,cols) // C = ones(rows,cols)
42 MatrixXd::Random(rows,cols) // rand(rows,cols)*2-1 // MatrixXd::Random returns uniform random numbers in (-1, 1).
43 C.setRandom(rows,cols) // C = rand(rows,cols)*2-1
44 VectorXd::LinSpaced(size,low,high) // linspace(low,high,size)'
45 v.setLinSpaced(size,low,high) // v = linspace(low,high,size)'
46 VectorXi::LinSpaced(((hi-low)/step)+1, // low:step:hi
47 low,low+step*(size-1)) //
50 // Matrix slicing and blocks. All expressions listed here are read/write.
51 // Templated size versions are faster. Note that Matlab is 1-based (a size N
52 // vector is x(1)...x(N)).
56 x.tail(n) // x(end - n + 1: end)
57 x.tail<n>() // x(end - n + 1: end)
58 x.segment(i, n) // x(i+1 : i+n)
59 x.segment<n>(i) // x(i+1 : i+n)
60 P.block(i, j, rows, cols) // P(i+1 : i+rows, j+1 : j+cols)
61 P.block<rows, cols>(i, j) // P(i+1 : i+rows, j+1 : j+cols)
64 P.leftCols<cols>() // P(:, 1:cols)
65 P.leftCols(cols) // P(:, 1:cols)
66 P.middleCols<cols>(j) // P(:, j+1:j+cols)
67 P.middleCols(j, cols) // P(:, j+1:j+cols)
68 P.rightCols<cols>() // P(:, end-cols+1:end)
69 P.rightCols(cols) // P(:, end-cols+1:end)
70 P.topRows<rows>() // P(1:rows, :)
71 P.topRows(rows) // P(1:rows, :)
72 P.middleRows<rows>(i) // P(i+1:i+rows, :)
73 P.middleRows(i, rows) // P(i+1:i+rows, :)
74 P.bottomRows<rows>() // P(end-rows+1:end, :)
75 P.bottomRows(rows) // P(end-rows+1:end, :)
76 P.topLeftCorner(rows, cols) // P(1:rows, 1:cols)
77 P.topRightCorner(rows, cols) // P(1:rows, end-cols+1:end)
78 P.bottomLeftCorner(rows, cols) // P(end-rows+1:end, 1:cols)
79 P.bottomRightCorner(rows, cols) // P(end-rows+1:end, end-cols+1:end)
80 P.topLeftCorner<rows,cols>() // P(1:rows, 1:cols)
81 P.topRightCorner<rows,cols>() // P(1:rows, end-cols+1:end)
82 P.bottomLeftCorner<rows,cols>() // P(end-rows+1:end, 1:cols)
83 P.bottomRightCorner<rows,cols>() // P(end-rows+1:end, end-cols+1:end)
85 // Of particular note is Eigen's swap function which is highly optimized.
87 R.row(i) = P.col(j); // R(i, :) = P(:, j)
88 R.col(j1).swap(mat1.col(j2)); // R(:, [j1 j2]) = R(:, [j2, j1])
90 // Views, transpose, etc;
93 R.transpose() // R.' or conj(R') // Read-write
94 R.diagonal() // diag(R) // Read-write
95 x.asDiagonal() // diag(x)
96 R.transpose().colwise().reverse() // rot90(R) // Read-write
97 R.rowwise().reverse() // fliplr(R)
98 R.colwise().reverse() // flipud(R)
99 R.replicate(i,j) // repmat(P,i,j)
102 // All the same as Matlab, but matlab doesn't have *= style operators.
103 // Matrix-vector. Matrix-matrix. Matrix-scalar.
104 y = M*x; R = P*Q; R = P*s;
105 a = b*M; R = P - Q; R = s*P;
106 a *= M; R = P + Q; R = P/s;
111 // Vectorized operations on each element independently
113 R = P.cwiseProduct(Q); // R = P .* Q
114 R = P.array() * s.array(); // R = P .* s
115 R = P.cwiseQuotient(Q); // R = P ./ Q
116 R = P.array() / Q.array(); // R = P ./ Q
117 R = P.array() + s.array(); // R = P + s
118 R = P.array() - s.array(); // R = P - s
119 R.array() += s; // R = R + s
120 R.array() -= s; // R = R - s
121 R.array() < Q.array(); // R < Q
122 R.array() <= Q.array(); // R <= Q
123 R.cwiseInverse(); // 1 ./ P
124 R.array().inverse(); // 1 ./ P
125 R.array().sin() // sin(P)
126 R.array().cos() // cos(P)
127 R.array().pow(s) // P .^ s
128 R.array().square() // P .^ 2
129 R.array().cube() // P .^ 3
130 R.cwiseSqrt() // sqrt(P)
131 R.array().sqrt() // sqrt(P)
132 R.array().exp() // exp(P)
133 R.array().log() // log(P)
134 R.cwiseMax(P) // max(R, P)
135 R.array().max(P.array()) // max(R, P)
136 R.cwiseMin(P) // min(R, P)
137 R.array().min(P.array()) // min(R, P)
138 R.cwiseAbs() // abs(P)
139 R.array().abs() // abs(P)
140 R.cwiseAbs2() // abs(P.^2)
141 R.array().abs2() // abs(P.^2)
142 (R.array() < s).select(P,Q ); // (R < s ? P : Q)
143 R = (Q.array()==0).select(P,A) // R(Q==0) = P(Q==0)
144 R = P.unaryExpr(ptr_fun(func)) // R = arrayfun(func, P) // with: scalar func(const scalar &x);
150 R.minCoeff() // min(R(:))
151 R.maxCoeff() // max(R(:))
152 s = R.minCoeff(&r, &c) // [s, i] = min(R(:)); [r, c] = ind2sub(size(R), i);
153 s = R.maxCoeff(&r, &c) // [s, i] = max(R(:)); [r, c] = ind2sub(size(R), i);
155 R.colwise().sum() // sum(R)
156 R.rowwise().sum() // sum(R, 2) or sum(R')'
157 R.prod() // prod(R(:))
158 R.colwise().prod() // prod(R)
159 R.rowwise().prod() // prod(R, 2) or prod(R')'
160 R.trace() // trace(R)
162 R.colwise().all() // all(R)
163 R.rowwise().all() // all(R, 2)
165 R.colwise().any() // any(R)
166 R.rowwise().any() // any(R, 2)
168 // Dot products, norms, etc.
170 x.norm() // norm(x). Note that norm(R) doesn't work in Eigen.
171 x.squaredNorm() // dot(x, x) Note the equivalence is not true for complex
172 x.dot(y) // dot(x, y)
173 x.cross(y) // cross(x, y) Requires #include <Eigen/Geometry>
177 A.cast<double>(); // double(A)
178 A.cast<float>(); // single(A)
179 A.cast<int>(); // int32(A)
182 // if the original type equals destination type, no work is done
184 // Note that for most operations Eigen requires all operands to have the same type:
185 MatrixXf F = MatrixXf::Zero(3,3);
186 A += F; // illegal in Eigen. In Matlab A = A+F is allowed
187 A += F.cast<double>(); // F converted to double and then added (generally, conversion happens on-the-fly)
189 // Eigen can map existing memory into Eigen matrices.
191 Vector3f::Map(array).fill(10); // create a temporary Map over array and sets entries to 10
192 int data[4] = {1, 2, 3, 4};
193 Matrix2i mat2x2(data); // copies data into mat2x2
194 Matrix2i::Map(data) = 2*mat2x2; // overwrite elements of data with 2*mat2x2
195 MatrixXi::Map(data, 2, 2) += mat2x2; // adds mat2x2 to elements of data (alternative syntax if size is not know at compile time)
197 // Solve Ax = b. Result stored in x. Matlab: x = A \ b.
198 x = A.ldlt().solve(b)); // A sym. p.s.d. #include <Eigen/Cholesky>
199 x = A.llt() .solve(b)); // A sym. p.d. #include <Eigen/Cholesky>
200 x = A.lu() .solve(b)); // Stable and fast. #include <Eigen/LU>
201 x = A.qr() .solve(b)); // No pivoting. #include <Eigen/QR>
202 x = A.svd() .solve(b)); // Stable, slowest. #include <Eigen/SVD>
203 // .ldlt() -> .matrixL() and .matrixD()
204 // .llt() -> .matrixL()
205 // .lu() -> .matrixL() and .matrixU()
206 // .qr() -> .matrixQ() and .matrixR()
207 // .svd() -> .matrixU(), .singularValues(), and .matrixV()
209 // Eigenvalue problems
211 A.eigenvalues(); // eig(A);
212 EigenSolver<Matrix3d> eig(A); // [vec val] = eig(A)
213 eig.eigenvalues(); // diag(val)
214 eig.eigenvectors(); // vec
215 // For self-adjoint matrices use SelfAdjointEigenSolver<>