线性代数的例子
本教程展示了一些使用 Wolfram 语言 进行涉及线性代数的计算的例子.
矩阵排序
某些稀疏矩阵技术尝试对矩阵重新排列,使元素组合成块. 然后利用高密度矩阵技术计算每个矩阵块. 一个简单的对矩阵按块排序的方式,涉及到按照每行元素的总和排序. 这将在这个例子中说明.
n = 30;
nz = 30;
sp = SparseArray[
RandomInteger[{1, n}, {nz, 2}] -> Table[1, {nz}], {n, n}];
sp = sp + Transpose[sp];MatrixPlot[sp]perm = Reverse[Ordering[sp.Table[1, {n}]]];
spOrdered = sp[[perm, perm]];MatrixPlot[spOrdered]n = 10000;
nz = 2000;
sp = SparseArray[
RandomInteger[{1, n}, {nz, 2}] -> Table[1, {nz}], {n, n}];
sp = sp + Transpose[sp];
perm = Reverse[Ordering[sp.Table[1, {n}]]];
spOrdered = sp[[perm, perm]];MatrixPlot[ sp]MatrixPlot[spOrdered]满秩最小二乘解
求解矩阵方程
是矩阵计算的关键问题之一;关于求解技巧在"求解线性系统"中有讨论. 如果
是一个
×
矩阵,这是由含有
个未知数的
. 如果
,即方程个数大与未知数个数,则方程组是超定系统. 一种用于找到最小二乘解的一般技巧在"最小二乘解集"中给出. 这个例子将展示一些不同的技术来找到满秩矩阵超定系统的近似解.
请注意对于本节所有例子中的解同样适用于在输入中给出稀疏矩阵的情形.
最小二乘 Cholesky
这个技巧采用 Cholesky 分解找到最小二乘解. 如果一个矩阵
有满秩,可以通过以下方式找到解.
L2CholeskySolve[a_, b_] :=
Module[ {aT, c, d, g, gT, y},
aT = Transpose[a];
c = aT.a;
d = aT.b;
g = CholeskyDecomposition[c];
gT = Transpose[g];
y = LinearSolve[gT, d];
LinearSolve[g, y]
]mat = {{1., 2.}, {5., 6.}, {4.5, 6.}};
vec = {5., 6., 8.};
sol1 = L2CholeskySolve[mat, vec]mat.sol1PseudoInverse[mat].vecMap[Norm[mat.sol1 - vec, #]&, {1, 2, Infinity}]最小二乘 QR
当
是满秩时,一种更准确的解决该问题的方法是使用如下的 QR 分解.
L2QRSolve[a_, b_] :=
Module[ {Q, R, m, n},
{m, n} = Dimensions[a];
{Q, R} = QRDecomposition[a];
LinearSolve[R, Q.b]
]mat = {{1., 2.}, {5., 6.}, {4.5, 6.}};
vec = {5., 6., 8.};
sol2 = L2QRSolve[mat, vec]mat = {{-4.1799231, -5.010860245276564, -5.841797048032377}, {4.985838406555408, 5.8047054724005, 6.623572538245592}, {3.6958074724742653, 3.687168314818264, 3.678529157162263}, {-5.488825789918355, -7.024200075119946, -8.559574360321538}};
vec = {1., 4., 2., 1.};
sol1 = L2CholeskySolve[mat, vec];
sol2 = L2QRSolve[mat, vec];Norm[mat.sol1 - vec]Norm[mat.sol2 - vec]1-范数和无穷-范数的最小化
当矩阵方程
超定(即
)时,找到近似解的许多技术要进行2范数最小化. 这是由于使得它在计算上易于处理的一定的优势. 原因之一是,函数
是可微的,并且微分是一个线性操作. 因此,可以构成一个找到最小解的线性系统. 另一个原因是,在正交变换下2范数可以保存. 这意味着可以构成一系列可能比较易于求解的等价问题.
然而,还可以通过最小化其它的范数,例如1范数或
范数,来找到另外的解. 在某些特定情况下,这些方式可能会由于找到的解能够保持一些相对于个别问题的重要性质,从而更可取. 在这个例子中展示了一些技术来找到使范数最小化的适当解; 两者都将使用一个方法来寻找约束线性问题的最小值;这通常被称为线性编程.
在 Wolfram 语言 中,线性编程由函数 LinearProgramming 实现. 它可以解决对 Wolfram 语言 所支持的不同类型的数值的线性编程问题:这些数值类型包括整数、有理数,以及机器精度近似实数和任意精度近似实数. 另外,它还通过内点法来提供适当的技术来使极大型系统最小化.
在本节中给出的解适用于密集矩阵的输入. 将这些解进行修改可直接作为稀疏矩阵输入,这将有必要充分利用内点法线性编程技术.
1-范数最小化
L1Solve[A_, b_] :=
Module[
{m, n, Aall, ball, c, x, bds, id},
{m, n} = Dimensions[A];
id = IdentityMatrix[m];
Aall = Join[Transpose[Join[Transpose[id], Transpose[-A]]], Transpose[Join[Transpose[id], Transpose[A]]]];
ball = Join[-b, b];
c = Join[Table[1, {m}], Table[0, {n}]];
bds = Table[{-Infinity, Infinity}, {m + n}];
x = LinearProgramming[c, Aall, ball, bds];
Take[x, -n]
]mat = {{1., 2.}, {5., 6.}, {4.5, 6.}};
vec = {5., 6., 8.};
sol3 = L1Solve[mat, vec]Map[Norm[mat.sol3 - vec, #]&, {1, 2, Infinity}]无穷-范数最小化
范数的最小化与1范数的最小化相似,涉及到找到使下式最小的
的值.
LInfinitySolve[A_, b_] :=
Module[
{m, n, Aall, ball, c, x, bds, onesT},
{m, n} = Dimensions[A];
onesT = {Table[1, {m}]};
Aall = Join[Transpose[Join[onesT, Transpose[-A]]], Transpose[Join[onesT, Transpose[A]]]];
ball = Join[-b, b];
c = Join[{1}, Table[0, {n}]];
bds = Table[{-Infinity, Infinity}, {n + 1}];
x = LinearProgramming[c, Aall, ball, bds];
Take[x, -n]
]mat = {{1., 2.}, {5., 6.}, {4.5, 6.}};
vec = {5., 6., 8.};
sol4 = LInfinitySolve[mat, vec]Map[Norm[mat.sol4 - vec, #]&, {1, 2, Infinity}]有限差分解
求解偏微分方程的一种方式是,形成一个空间离散并求其有限差分解. 这个例子考察一个单位正方形上的泊松方程的基本有限差分解.
num = 80;hStep = N[1 / (num + 1)];
vars = Table[u[i, j], {i, num}, {j, num}];
Short[vars, 6]kern = {{0, 1, 0}, {1, -4, 1}, {0, 1, 0}} / hStep ^ 2;lap = ListCorrelate[kern, vars, {2, 2}, 0];Short[eqns = Thread[Map[Flatten, lap == Table[Exp[-((hStep i)^2 + (hStep j)^2)], {i, num}, {j, num}]]], 20]{vec, mat} = CoefficientArrays[eqns, Flatten[vars]]sol = LinearSolve[mat, -vec];//Timingsol = Partition[sol, num];ListContourPlot[sol, ColorFunction -> Hue]factors = LinearSolve[mat]cvec = Flatten[Table[Cos[6(hStep i)^2 + 3(hStep j)^2], {i, num}, {j, num}]];
sol = Partition[factors[cvec], num];ListContourPlot[sol, ColorFunction -> Hue]sol = PadRight[PadLeft[sol, {num + 1, num + 1}], {num + 2, num + 2}];
data = Flatten[Table[{i hStep, j hStep, sol[[i + 1, j + 1]]}, {j, 0, num + 1}, {i, 0, num + 1}], 1];
iFun = Interpolation[data]Plot3D[iFun[x, y], {x, 0, 1}, {y, 0, 1}]Table[iFun[x, 0.5], {x, 0, 1, 0.05}]网格分区
这个例子演示了一个极端特征值的实际用途. 其目的是将网格或图形的顶点分割为两个子域,各自含有数量大致相等的顶点,分区贯穿尽可能少的边缘.(边缘在当它的两个顶点分别在每个子域上被削去.)对图形分区的一种方式是形成拉普拉斯并使用其费德勒向量(对应第二最小特征值的特征向量)以对顶点排序. 对图形分区还存在更高效的算法,不需要进行特征向量的计算,但是费德勒方法仍然是重要的一种方法.
数据
coordinates = {{-60., 60.}, {-60., -60.}, {60., 60.}, {60., -60.}, {12.809475, 12.5}, {-12.809475, -12.5}, {-7., -4.}, {-7., 2.}, {-60., 0.}, {0., 60.}, {60., 0.}, {0., -60.}, {-1., 7.415715}, {1., -7.415715}, {-7., -1.}, {9., 8.716814}, {-1., -13.474961}, {4.769961, 25.682898}, {1., 13.474961}, {-4.769961, -25.682898}, {2.736733, -23.013988}, {-8.206816, -29.954422}, {5., 1.037611}, {5., 13.301991}, {8.822488, 14.823734}, {-7.494958, -19.97955}, {-2.746936, -18.232192}, {13.830102, 22.165298}, {7.494958, 19.97955}, {2.746936, 18.232192}, {-13.830102, -22.165298}, {-17.681008, -21.469559}, {-7.905112, -27.965817}, {-22.988602, -13.396162}, {8.897511, -15.808074}, {6.245633, 15.852277}, {-11.891051, -17.475286}, {11.891052, 17.475286}, {-17.238035, -9.740396}, {-1., -3.029623}, {-1., 17.861054}, {1., 3.029623}, {-1., 2.193046}, {1., -2.193046}, {1., -17.861054}, {1., -12.638385}, {-1., -8.252292}, {1., 8.252292}, {-0.561635, -22.748417}, {-9.895431, -25.457646}, {-1.937437, -26.879819}, {5., -11.226769}, {4.923133, -17.356056}, {3.197251, -26.32948}, {0.647938, -25.24711}, {-2.153142, -29.248212}, {-18.935854, -24.655299}, {5., 7.169801}, {-14.955767, -17.159894}, {-18.991451, -15.547121}, {-13.641527, -25.994504}, {-15.95189, -23.991475}, {-10.89993, -27.348592}, {-4.851524, -27.806102}, {-18.098166, -2.47478}, {-60., -30.}, {-39.049083, 28.76261}, {-39.467927, -12.32765}, {-39.049083, -1.23739}, {-30., -60.}, {-30.5, 38.930527}, {-9.467927, -42.32765}, {-9.549083, 7.693137}, {-4.103408, -44.977211}, {-4., 9.930527}, {1.598625, -43.16474}, {-0.5, 38.930527}, {6.915051, 41.082649}, {10.853493, -1.654037}, {31.598625, -43.16474}, {34.448756, -37.904037}, {30., 60.}, {34.448756, -7.904037}, {36.915051, 11.082649}, {60., 30.}, {-20.024542, 23.311832}, {21.915051, 41.082649}, {36.915051, 41.082649}, {-60., -15.}, {-60., 30.}, {-39.467927, -42.32765}, {-49.524542, 14.381305}, {-49.524542, 44.381305}, {-41.494301, -6.698081}, {-39.467927, -27.32765}, {-28.573625, 13.143915}, {-30., 60.}, {-29.536812, 26.037221}, {-29.20189, -18.491474}, {-28.573625, -1.856085}, {-24.467927, -42.32765}, {-15.75, 28.39579}, {-19.061354, 10.418526}, {-15., -60.}, {-14.20189, -33.491474}, {-15.25, 49.465263}, {-10.512271, 20.586443}, {-6.155112, -37.465817}, {-3.128275, -37.112712}, {-4.733963, -51.163825}, {-1.076571, -44.624106}, {2.397938, -34.74711}, {-0.75, 28.39579}, {2.384981, 42.841449}, {10.372577, 31.623973}, {9., -3.547565}, {10.957526, 50.541324}, {15.799313, -51.58237}, {17.397938, -34.74711}, {17.872577, 31.623973}, {21.673134, -26.856056}, {15., 60.}, {21.673134, -11.856056}, {23.629115, 2.297981}, {25.372577, 16.623973}, {18.823003, -32.116759}, {25.372577, 31.623973}, {25.957526, 50.541324}, {30., -60.}, {36.404738, 6.25}, {36.915051, 26.082649}, {47.224378, -48.952019}, {45., 60.}, {47.224378, -18.952019}, {60., -30.}, {60., 45.}, {60., 15.}, {-60., 15.}, {-60., 45.}, {-60., -45.}, {-49.733963, -21.163825}, {-49.733963, -6.163825}, {-49.524542, 29.381305}, {-49.524542, -0.618695}, {-45.25, 49.465263}, {-49.733963, -36.163825}, {-45., -60.}, {-39.049083, 13.76261}, {-40.012271, 41.655916}, {-32.241451, -10.047122}, {-31.018842, -7.316776}, {-29.20189, -33.491474}, {-33.811354, 20.953263}, {-33.811354, 5.953263}, {-34.733963, -51.163825}, {-31.228264, -12.861906}, {-29.20189, -25.991474}, {-25.262271, 31.121179}, {-30.25, 49.465263}, {-23.335896, 5.334568}, {-19.733963, -51.163825}, {-21.70189, -33.491474}, {-14.786812, 15.502485}, {-15.5, 38.930527}, {-15., 60.}, {-12.233963, -51.163825}, {-8.375, 23.128422}, {-8.125, 33.663159}, {-0.25, 49.465263}, {-2.051704, -52.488606}, {-1.614857, -36.936159}, {0.799313, -51.58237}, {3.577471, 34.262174}, {3.457526, 50.541324}, {7.899656, -55.791185}, {9.498282, -38.955925}, {5.842506, 33.382774}, {10.297594, -30.538295}, {12.435192, -26.592768}, {11.010127, -29.223119}, {15., -60.}, {18.457526, 50.541324}, {16.598625, -43.16474}, {15.285322, -21.332065}, {22.651124, -4.779037}, {25.210814, -37.640749}, {24.607106, 9.375}, {28.060945, -32.380046}, {24.862263, 11.791324}, {25.372577, 24.123973}, {23.698969, -47.373555}, {28.060945, -17.380046}, {34.448756, -22.904037}, {33.457526, 50.541324}, {40.957526, 50.541324}, {40.836567, -43.428028}, {45.799313, -51.58237}, {47.224378, -3.952019}, {48.457526, 5.541324}, {48.457526, 20.541324}, {48.457526, 35.541324}, {47.224378, -33.952019}, {48.457526, 50.541324}, {60., -15.}, {60., -45.}, {-54.762271, 37.190653}, {-54.762271, 52.190653}, {-54.762271, 7.190653}, {-54.762271, 22.190653}, {-49.733963, -51.163825}, {-50.74715, -3.349041}, {-49.733963, -43.663825}, {-49.733963, -28.663825}, {-49.733963, -13.663825}, {-44.286812, 6.571958}, {-44.286812, 36.571958}, {-44.286812, 21.571958}, {-45., 60.}, {-42.233963, -51.163825}, {-37.875, 44.197895}, {-39.049083, 6.26261}, {-37.625, 54.732632}, {-31.967927, -42.32765}, {-27.233963, -51.163825}, {-22.875, 44.197895}, {-22.625, 54.732632}, {-23.125, 33.663159}, {-7.875, 44.197895}, {-7.625, 54.7 ... 379}, {45., -60.}, {48.202369, 3.125}, {48.457526, 43.041324}, {48.457526, 13.041324}, {48.457526, 28.041324}, {53.612189, -54.476009}, {53.612189, -9.476009}, {53.612189, -39.476009}, {53.612189, -24.476009}, {-52.625, 54.732632}, {38.698969, -47.373555}, {52.899656, -55.791185}, {-39.467927, -34.82765}, {-39.467927, -19.82765}, {-33.811354, -1.546737}, {-34.334909, -30.409562}, {-28.573625, 5.643915}, {-27.108433, -13.129034}, {-25.781113, -7.626123}, {-24.068872, -21.573387}, {-24.068872, -29.073387}, {-26.834909, -37.909562}, {-23.335896, -2.165432}, {-18.57976, 3.971873}, {-19.334909, -37.909562}, {-16.568872, -29.073387}, {-18.716608, -18.592506}, {-15.900463, -14.023561}, {-16.967927, -42.32765}, {-15.755555, -21.817429}, {-13.823625, 2.609179}, {-14.805062, -19.755279}, {-13.735814, -24.079901}, {-14.085381, -28.34564}, {-12.133847, -24.135439}, {-12.868763, -5.533022}, {-8.837372, -36.141036}, {-8.25, 28.39579}, {-8., 38.930527}, {-7.351243, -25.992286}, {-5.274542, 12.777095}, {-6.904738, -10.376146}, {-6.904738, -12.987481}, {-4., 0.596523}, {-4., -2.014812}, {-5.667546, -33.533567}, {-0.875, 23.128422}, {-0.625, 33.663159}, {-4., -6.126146}, {-2.424183, -24.562157}, {-1.249536, -24.814118}, {-0.277258, -36.206476}, {1.087549, -22.881202}, {1.868367, -20.437521}, {2.797594, -30.538295}, {2.134981, 32.306713}, {2.981226, 38.551811}, {2.961567, -14.99722}, {4.193324, -20.325447}, {3.873468, 15.767091}, {5.548454, -41.060333}, {6.347766, -32.642703}, {4.927766, 19.630794}, {7.571269, 28.653436}, {8.158723, 17.401642}, {7.068019, -21.621736}, {7.816221, -26.461124}, {9.875502, -8.731056}, {12.101339, 26.894636}, {9.897938, -34.74711}, {9., 11.782909}, {11.831484, 5.422981}, {12.900312, 15.089181}, {12.393814, 36.353311}, {10.666352, -21.200421}, {14.415051, 41.082649}, {16.263313, -6.755046}, {15.851339, 26.894636}, {15.285322, -13.832065}, {18.219295, 7.398991}, {19.091026, 14.561987}, {19.601339, 26.894636}, {17.241304, 0.321972}, {21.673134, -19.356056}, {18.708291, 10.9375}, {19.601339, 19.394636}, {19.893814, 36.353311}, {29.415051, 41.082649}, {30.016926, 4.273991}, {31.143814, 13.853311}, {31.143814, 28.853311}, {31.143814, 21.353311}, {29.038936, -2.803028}, {36.915051, 18.582649}, {36.915051, 33.582649}, {-27.615026, -11.721642}, {9.098625, -43.16474}, {-27.003722, -10.356469}, {-34.334909, -22.909562}, {-34.334909, -15.409562}, {-29.20189, -22.241474}, {-23.528737, -17.484774}, {-25.95476, 1.739241}, {-24.068872, -25.323387}, {-24.299083, 18.227874}, {-21.93964, -5.050452}, {-21.602822, -7.46107}, {-22.054429, -19.418871}, {-20.717031, 1.429894}, {-20.318872, -29.073387}, {-17.95189, -33.491474}, {-18.48769, -12.531142}, {-16.318387, -19.314727}, {-15.960896, 0.067199}, {-15.023755, -11.120198}, {-12.839119, -9.016511}, {-14.305219, 9.055832}, {-11.834909, -37.909562}, {-15.625, 33.663159}, {-12.983163, 0.839197}, {-11.204353, -31.722948}, {-11.768479, -25.726075}, {-9.904737, -8.25}, {-10.152217, -16.239775}, {-8.900272, -26.711731}, {-9.857106, -11.438073}, {-10.030677, 14.13979}, {-7.778206, -15.366096}, {-5.5, 5.965263}, {-6.952369, -7.188073}, {-4.5625, 25.762106}, {-4.4375, 31.029475}, {-3.952369, -11.925554}, {-4., -3.514812}, {-3.952369, -9.314219}, {-3.353699, -26.281359}, {-4.825837, -15.609836}, {-4.6875, 20.494738}, {-4.3125, 36.296843}, {-4., 4.707857}, {-0.375, 44.197895}, {-1., -5.640958}, {0.620987, -20.433219}, {0.561635, 22.748417}, {1.947481, 24.40566}, {3.452471, 28.994805}, {2.596662, -19.093251}, {1.359079, 16.044508}, {2.961567, -17.608555}, {3.980783, -13.111995}, {3., 13.388476}, {4.173716, 29.972536}, {4.496285, 17.042234}, {5., -14.292864}, {5.171805, 14.810988}, {5.358865, -24.161617}, {5.306234, 29.532836}, {6.870296, 17.915913}, {6.747423, -28.433888}, {8.936288, 45.811987}, {8.643814, 36.353311}, {7.741415, 15.8575}, {6.706887, 31.018105}, {8.43565, 26.288767}, {9.751605, -24.107252}, {9.969867, 19.213434}, {10.35677, 16.14951}, {11.236958, 29.259304}, {10.815981, 13.661867}, {9.895432, 25.457646}, {10.904737, 10.608407}, {9.926747, 3.531389}, {10.8614, 14.956458}, {12.32048, 8.961491}, {12.580412, -11.28156}, {13.111958, 29.259304}, {12.091417, -18.57007}, {14.122577, 31.623973}, {14.536394, 2.872477}, {14.840721, 24.529967}, {15.514385, 9.949495}, {16.143814, 36.353311}, {14.955767, 17.159894}, {16.436288, 45.811987}, {16.715721, 24.529967}, {17.054163, -26.724412}, {16.752309, -3.216537}, {16.460564, 18.363642}, {18.968223, -9.305551}, {16.861958, 29.259304}, {18.835869, 12.145662}, {19.601339, 23.144636}, {20.924205, 4.848486}, {21.413201, 8.386995}, {21.622577, 31.623973}, {22.231801, 15.59298}, {25.372577, 27.873973}, {25.372577, 20.373973}, {27.393814, 36.353311}, {23.936288, 45.811987}, {27.312016, 6.824495}, {28.060945, -24.880046}, {30.505922, 7.8125}, {31.143814, 36.353311}, {31.143814, 17.603311}, {31.143814, 25.103311}, {30.761079, 10.228824}, {41.814558, 1.148991}, {-5.756135, 19.223748}, {3., 10.777142}, {7.103689, -27.7763}, {9., 2.584625}};triangles = {{209, 90, 138}, {206, 139, 90}, {207, 1, 139}, {208, 138, 9}, {142, 9, 89}, {141, 89, 66}, {210, 140, 2}, {146, 66, 140}, {250, 1, 207}, {206, 93, 139}, {207, 139, 93}, {208, 92, 138}, {209, 138, 92}, {208, 9, 144}, {206, 90, 143}, {209, 143, 90}, {211, 9, 142}, {211, 144, 9}, {212, 146, 140}, {213, 141, 66}, {213, 66, 146}, {214, 142, 89}, {214, 89, 141}, {212, 140, 210}, {250, 218, 1}, {210, 2, 147}, {250, 207, 145}, {211, 142, 144}, {207, 93, 145}, {215, 92, 208}, {144, 215, 208}, {216, 93, 206}, {143, 216, 206}, {209, 92, 217}, {143, 209, 217}, {250, 145, 218}, {210, 219, 212}, {142, 214, 94}, {142, 94, 144}, {146, 253, 213}, {141, 254, 214}, {212, 91, 146}, {213, 95, 141}, {219, 210, 147}, {149, 145, 93}, {216, 149, 93}, {215, 144, 69}, {216, 143, 67}, {217, 67, 143}, {215, 148, 92}, {217, 92, 148}, {219, 91, 212}, {214, 68, 94}, {144, 94, 69}, {253, 95, 213}, {254, 68, 214}, {254, 141, 95}, {253, 146, 91}, {222, 218, 145}, {216, 67, 149}, {220, 145, 149}, {221, 215, 69}, {221, 148, 215}, {219, 147, 155}, {145, 220, 222}, {153, 67, 217}, {148, 153, 217}, {219, 155, 91}, {255, 69, 94}, {256, 95, 253}, {339, 254, 95}, {340, 68, 254}, {150, 94, 68}, {222, 97, 218}, {221, 154, 148}, {221, 69, 154}, {253, 91, 223}, {155, 147, 70}, {149, 67, 71}, {220, 149, 71}, {254, 339, 340}, {339, 95, 256}, {255, 154, 69}, {151, 255, 94}, {223, 91, 155}, {150, 68, 340}, {220, 159, 222}, {150, 151, 94}, {256, 253, 152}, {153, 98, 67}, {154, 96, 148}, {153, 148, 96}, {152, 253, 223}, {98, 71, 67}, {220, 71, 159}, {222, 159, 97}, {339, 99, 340}, {157, 339, 256}, {156, 150, 340}, {154, 255, 257}, {340, 99, 156}, {155, 224, 223}, {255, 151, 100}, {256, 152, 157}, {341, 339, 157}, {341, 99, 339}, {224, 155, 70}, {336, 150, 156}, {257, 96, 154}, {255, 100, 257}, {338, 151, 150}, {262, 152, 223}, {345, 98, 153}, {258, 156, 99}, {336, 338, 150}, {96, 345, 153}, {258, 336, 156}, {259, 100, 151}, {158, 71, 98}, {338, 259, 151}, {224, 101, 223}, {225, 159, 71}, {262, 223, 101}, {343, 257, 100}, {226, 97, 159}, {157, 344, 341}, {261, 157, 152}, {341, 260, 99}, {338, 336, 258}, {257, 160, 96}, {258, 99, 342}, {227, 71, 158}, {343, 100, 263}, {343, 160, 257}, {262, 162, 152}, {263, 100, 259}, {344, 260, 341}, {344, 157, 261}, {34, 338, 258}, {224, 70, 161}, {342, 99, 260}, {225, 71, 227}, {338, 34, 259}, {159, 225, 226}, {261, 152, 162}, {158, 98, 86}, {345, 86, 98}, {34, 258, 342}, {345, 96, 103}, {224, 161, 101}, {346, 263, 259}, {160, 103, 96}, {262, 101, 265}, {347, 259, 34}, {343, 263, 349}, {160, 343, 349}, {342, 260, 348}, {346, 259, 347}, {342, 348, 34}, {261, 350, 344}, {162, 262, 265}, {226, 165, 97}, {344, 57, 260}, {350, 261, 162}, {348, 260, 57}, {161, 70, 104}, {227, 158, 102}, {346, 65, 263}, {350, 57, 344}, {349, 264, 160}, {349, 263, 65}, {346, 347, 65}, {164, 225, 227}, {269, 101, 161}, {158, 86, 102}, {264, 103, 160}, {265, 101, 269}, {225, 106, 226}, {162, 351, 350}, {86, 345, 163}, {351, 162, 265}, {345, 103, 163}, {350, 266, 57}, {264, 349, 354}, {351, 266, 350}, {349, 65, 354}, {359, 227, 102}, {359, 164, 227}, {225, 164, 106}, {352, 60, 268}, {226, 106, 165}, {353, 267, 32}, {267, 59, 60}, {357, 103, 264}, {351, 265, 105}, {355, 39, 352}, {353, 59, 267}, {268, 60, 59}, {353, 32, 270}, {266, 274, 57}, {268, 355, 352}, {270, 32, 62}, {161, 166, 269}, {351, 105, 266}, {354, 271, 264}, {357, 163, 103}, {269, 358, 265}, {354, 65, 360}, {166, 161, 104}, {272, 353, 270}, {271, 357, 264}, {107, 102, 86}, {353, 272, 59}, {270, 62, 31}, {358, 105, 265}, {163, 107, 86}, {355, 356, 39}, {266, 105, 274}, {270, 31, 272}, {355, 268, 6}, {268, 59, 6}, {273, 31, 62}, {273, 62, 61}, {356, 276, 39}, {354, 360, 271}, {355, 6, 356}, {273, 275, 31}, {359, 102, 278}, {273, 61, 275}, {361, 274, 105}, {359, 168, 164}, {163, 357, 367}, {166, 72, 269}, {228, 106, 164}, {358, 269, 72}, {229, 165, 106}, {357, 271, 73}, {362, 275, 61}, {362, 61, 63}, {356, 363, 276}, {6, 366, 356}, {367, 107, 163}, {358, 277, 105}, {364, 6, 37}, {167, 102, 107}, {361, 105, 277}, {357, 73, 367}, {360, 8, 271}, {362, 50, 275}, {361, 22, 274}, {6, 364, 366}, {366, 363, 356}, {362, 63, 50}, {278, 102, 167}, {168, 359, 278}, {166, 104, 110}, {279, 164, 168}, {279, 228, 164}, {229, 106, 228}, {271, 8, 73}, {358, 72, 277}, {363, 7, 276}, {365, 50, 63}, {364, 37, 26}, {364, 368, 366}, {365, 63, 33}, {366, 282, 363}, {449, 107, 367}, {365, 280, 50}, {72, 166, 74}, {277, 286, 361}, {364, 26, 368}, {286, 22, 361}, {367, 73, 281}, {449, 167, 107}, {283, 366, 368}, {277, 72, 108}, {365, 33, 280}, {370, 7, 363}, {370, 363, 282}, {366, 283, 282}, {229, 10, 165}, {369, 73, 8}, {371, 278, 167}, {166, 110, 74}, {281, 449, 367}, {372, 168, 278}, {277, 108, 286}, {379, 279, 168}, {280, 33, 64}, {377, 368, 26}, {110, 104, 12}, {108, 72, 74}, {377, 283, 368}, {369, 75, 73}, {281, 73, 75}, {378, 167, 449}, {285, 15, 7}, {284, 8, 15}, {370, 289, 7}, {282, 375, 370}, {373, 282, 283}, {167, 378, 371}, {372, 278, 371}, {20, 280, 64}, {168, 372, 379}, {380, 369, 8}, {279 ... 407}, {302, 310, 178}, {315, 35, 306}, {402, 25, 407}, {304, 404, 408}, {405, 315, 306}, {308, 116, 35}, {412, 16, 452}, {401, 314, 78}, {451, 178, 180}, {305, 407, 406}, {409, 25, 311}, {452, 116, 79}, {411, 311, 16}, {408, 115, 304}, {408, 404, 410}, {412, 452, 79}, {405, 307, 179}, {407, 25, 413}, {400, 78, 316}, {413, 25, 409}, {311, 411, 409}, {412, 312, 16}, {307, 180, 179}, {415, 308, 35}, {401, 115, 314}, {417, 35, 315}, {407, 38, 406}, {411, 16, 414}, {232, 175, 118}, {405, 179, 315}, {407, 413, 38}, {414, 16, 312}, {309, 408, 410}, {314, 316, 78}, {337, 232, 183}, {411, 5, 409}, {413, 409, 313}, {408, 416, 115}, {308, 317, 116}, {337, 183, 176}, {412, 79, 419}, {413, 313, 38}, {411, 414, 5}, {317, 79, 116}, {417, 319, 35}, {312, 412, 419}, {117, 400, 424}, {408, 309, 416}, {409, 5, 313}, {415, 35, 319}, {310, 176, 119}, {309, 410, 420}, {418, 314, 115}, {310, 119, 178}, {416, 418, 115}, {417, 315, 184}, {315, 179, 184}, {410, 28, 420}, {175, 181, 118}, {178, 119, 180}, {415, 317, 308}, {414, 312, 421}, {400, 316, 424}, {426, 179, 180}, {414, 421, 5}, {416, 309, 318}, {232, 118, 183}, {419, 79, 323}, {422, 314, 418}, {417, 184, 319}, {420, 318, 309}, {422, 316, 314}, {176, 183, 119}, {427, 79, 317}, {430, 418, 416}, {182, 122, 117}, {419, 320, 312}, {426, 184, 179}, {427, 323, 79}, {428, 28, 423}, {420, 28, 425}, {421, 312, 320}, {318, 430, 416}, {424, 182, 117}, {415, 319, 429}, {423, 5, 321}, {126, 426, 180}, {28, 428, 425}, {431, 5, 421}, {180, 119, 126}, {420, 425, 318}, {317, 415, 429}, {120, 422, 418}, {430, 120, 418}, {422, 327, 316}, {428, 423, 321}, {431, 321, 5}, {425, 322, 318}, {319, 184, 324}, {430, 318, 322}, {421, 320, 325}, {419, 323, 433}, {424, 316, 87}, {428, 432, 425}, {325, 431, 421}, {320, 419, 433}, {233, 118, 181}, {422, 120, 327}, {324, 184, 426}, {430, 322, 120}, {428, 321, 326}, {326, 432, 428}, {427, 317, 185}, {425, 432, 322}, {429, 319, 123}, {191, 183, 118}, {327, 87, 316}, {427, 185, 323}, {426, 126, 121}, {429, 185, 317}, {434, 325, 320}, {234, 119, 183}, {123, 319, 324}, {182, 424, 440}, {431, 325, 434}, {435, 120, 322}, {435, 327, 120}, {182, 128, 122}, {436, 321, 431}, {121, 324, 426}, {434, 320, 433}, {436, 326, 321}, {186, 126, 119}, {433, 323, 124}, {424, 87, 440}, {233, 191, 118}, {429, 123, 185}, {323, 185, 124}, {432, 326, 438}, {432, 190, 322}, {183, 191, 234}, {189, 431, 434}, {189, 436, 431}, {435, 322, 437}, {435, 127, 327}, {434, 433, 187}, {234, 186, 119}, {438, 326, 436}, {233, 181, 129}, {440, 128, 182}, {188, 121, 126}, {439, 87, 327}, {190, 432, 438}, {437, 322, 190}, {235, 126, 186}, {187, 189, 434}, {128, 82, 122}, {121, 442, 324}, {324, 192, 123}, {124, 441, 433}, {127, 435, 437}, {436, 189, 125}, {127, 439, 327}, {441, 187, 433}, {125, 438, 436}, {185, 123, 333}, {235, 188, 126}, {440, 87, 328}, {333, 124, 185}, {236, 191, 233}, {442, 192, 324}, {442, 121, 188}, {439, 328, 87}, {440, 328, 128}, {191, 80, 234}, {447, 189, 187}, {441, 124, 329}, {234, 80, 186}, {330, 125, 189}, {190, 446, 437}, {125, 445, 438}, {438, 332, 190}, {437, 331, 127}, {187, 441, 447}, {329, 124, 333}, {129, 236, 233}, {439, 127, 444}, {192, 83, 123}, {123, 83, 333}, {236, 80, 191}, {235, 186, 81}, {189, 447, 330}, {446, 331, 437}, {127, 331, 444}, {445, 332, 438}, {445, 125, 330}, {446, 190, 332}, {441, 329, 443}, {439, 444, 328}, {441, 443, 447}, {128, 328, 194}, {235, 81, 188}, {194, 82, 128}, {442, 188, 193}, {442, 193, 192}, {186, 80, 81}, {329, 333, 130}, {443, 329, 130}, {444, 88, 328}, {443, 130, 447}, {236, 129, 237}, {447, 84, 330}, {331, 335, 444}, {445, 334, 332}, {446, 131, 331}, {445, 330, 334}, {446, 332, 131}, {328, 88, 194}, {251, 80, 236}, {239, 83, 192}, {193, 239, 192}, {188, 81, 238}, {193, 188, 238}, {447, 130, 84}, {194, 195, 82}, {331, 131, 335}, {332, 334, 131}, {330, 84, 334}, {444, 335, 88}, {333, 83, 448}, {196, 81, 80}, {130, 333, 448}, {237, 251, 236}, {251, 240, 80}, {194, 88, 195}, {240, 196, 80}, {237, 129, 241}, {195, 133, 82}, {199, 84, 130}, {334, 84, 244}, {131, 245, 335}, {335, 201, 88}, {334, 200, 131}, {251, 237, 197}, {239, 193, 134}, {238, 81, 202}, {196, 202, 81}, {238, 134, 193}, {239, 198, 83}, {448, 83, 198}, {251, 197, 240}, {195, 88, 243}, {130, 448, 199}, {240, 132, 196}, {237, 241, 197}, {240, 197, 132}, {200, 334, 244}, {245, 201, 335}, {243, 88, 201}, {244, 84, 199}, {245, 131, 200}, {195, 203, 133}, {448, 198, 242}, {203, 195, 243}, {448, 242, 199}, {238, 202, 249}, {134, 238, 249}, {247, 198, 239}, {134, 247, 239}, {248, 202, 196}, {132, 248, 196}, {252, 197, 241}, {246, 132, 197}, {252, 246, 197}, {203, 3, 133}, {242, 198, 11}, {242, 11, 199}, {243, 201, 136}, {243, 136, 203}, {244, 199, 137}, {244, 137, 200}, {245, 85, 201}, {245, 200, 85}, {252, 241, 4}, {247, 11, 198}, {248, 135, 202}, {246, 205, 132}, {249, 204, 134}, {249, 202, 135}, {248, 132, 205}, {247, 134, 204}, {252, 4, 246}, {201, 85, 136}, {203, 136, 3}, {199, 11, 137}, {200, 137, 85}, {248, 205, 135}, {246, 4, 205}, {247, 204, 11}, {249, 135, 204}};绘制网格
edges = {};
FaceToEdges[{j_, k_, l_}] := { {j, k}, {k, j}, {j, l}, {l, j}, {k, l}, {l, k}};
edges = Partition[Flatten[Map[(FaceToEdges[#])&, triangles]], {2}];
graphicsEdge[{i_, j_}] := Line[{coordinates[[i]], coordinates[[j]]}];
Lines = Map[graphicsEdge, edges];
Show[Graphics[Lines], AspectRatio -> 1]拉普拉斯算符
一幅图形的拉普拉斯算符,
,是一个
×
离散矩阵,在图形中有
个顶点.
的输入值大多为零,对角线是与顶点次数对应的正整数,这里次数指的是与该顶点相连的顶点的个数. 另外,如果两个顶点,
和
通过一条边相连,则
为 -1.
num = Max[edges];
mat = SparseArray[Map[Rule[#, -1]&, edges], {num, num}]diag = SparseArray[{{i_, i_} -> 1.}, {num, num}] SparseArray[ -mat.Table[1., {num}]]lap = mat + diagMatrixPlot[lap]费德勒向量
每一行之和为零,因此拉普拉斯算符矩阵是具有零特征值的正半定矩阵. 对应于第二最小特征值的特征向量称为费德勒向量.
{{lam2, lam1}, {vec2, vec1}} = Eigensystem[lap, -2];Chop[{lam1, lam2}]节点的分区
order = Ordering[vec2];
domain1 = Drop[order, Floor[num / 2]];
domain2 = Drop[order, -(num - Floor[num / 2])];Map[(NodeColor[#] = Hue[0.])&, domain1];
Map[(NodeColor[#] = Hue[0.75])&, domain2];graphicsEdge[{i_, j_}] := {NodeColor[i], Line[{coordinates[[i]], (coordinates[[j]] + coordinates[[i]]) / 2.}],
NodeColor[j], Line[{coordinates[[j]], (coordinates[[j]] + coordinates[[i]]) / 2.}]
};
Lines = Flatten[Map[graphicsEdge, edges]];
Show[Graphics[Lines], AspectRatio -> 1]vec = Table[1, {num}];
vec[[domain1]] = 0;
Total[Abs[lap.vec]] / 2带有 NDSolve 的矩阵函数
Wolfram 语言 的诸多数值求解程序的重要特征之一是它们适用于向量和矩阵输入. 通常这用于求解耦合方程组,但也可用于解决矩阵特有的问题. 这里将给出一个使用数值微分方程求解程序 NDSolve 来找到矩阵函数的参数化解的例子.
matA = N[-{{1, 2, 3}, {4, 5, 6}, {7, 8, 9}}];
matA//MatrixFormDSolve[{x'[t] == a x[t], x[0] == 1}, x, t]solExp = NDSolve[X'[t] == matA.X[t] && X[0] == IdentityMatrix[3], X, {t, 0, 5}];
matrixExpA = X /. First[solExp]matrixExpA[0.]//MatrixFormexpA = matrixExpA[1.];
expA.expA - matrixExpA[2.]//MatrixFormdat = Table[{t, Norm[MatrixExp[N[matA], "Parameter" -> t] - matrixExpA[t], "Frobenius"]}, {t, 0.0001, 5, 0.2}];
ListLogPlot[dat]DSolve[{ x''[t] + a^2 x[t] == 0, x[0] == 0, x'[0] == a}, x, t]solSin = NDSolve[X''[t] + matA.matA.X[t] == 0 && X[0] == DiagonalMatrix[{0, 0, 0}] && X'[0] == matA, X, {t, 0, 4Pi}];
matrixSinA = X /. First[solSin]matrixSinA[0.]//MatrixFormsolCos = NDSolve[X''[t] + matA.matA.X[t] == 0 && X[0] == IdentityMatrix[3] && X'[0] == DiagonalMatrix[{0, 0, 0}], X, {t, 0, 4Pi}];
matrixCosA = X /. First[solCos]matrixCosA[0.]//MatrixForm(mat1 = matrixCosA[3.];mat1.mat1) + (mat2 = matrixSinA[3.];mat2.mat2)//MatrixForm相关技术笔记
-
▪
- Wolfram 语言中的线性代数 ▪
- 矩阵与张量运算 ▪
- 稀疏数组的运用 ▪
- 矩阵计算 ▪
- 矩阵类型