# NDSolve matrix of differential equations

If I have a system of differential equations that I can write in matrix form

$$\frac{\text{d}\mathbf{X}(t)}{\text{d}t} = \mathbf{A}\cdot \mathbf{X}(t),$$

with known initial conditions ($$\mathbf{X}(0)=\mathbf{X}_0)$$, how can I simulate the evolution of the components of $$\mathbf{X}(t)$$ using NDSolve?

• Note that the particular ODE you presented can be easily solved with MatrixExp[]. Sep 26 '20 at 11:46

---UPDATE: A simpler approach---

As per the comment by @J. M.'s discontentment, we can use the following code. Here, random numbers are used for A and X0 for illustration.

ClearAll["Global*"];
tEnd = 10;
X0 = RandomReal[{-1, 1}, 4];
A = RandomReal[{-1, 1}, {4, 4}];
sol = x[t] /.
NDSolve[{x'[t] == A.x[t], x == X0}, x, {t, 0, tEnd}];
Plot[sol, {t, 0, tEnd}]


---ORIGINAL ANSWER: Unnecessarily complicated approach for most (if not all) cases---

In the code below, the 'user' specifies the following:

• L: The size of $$\mathbf{X}$$, i.e. the number of simultaneous equations
• tEnd: The simulation duration
• X0: The initial conditions $$\mathbf{X}_0$$ (here a random vector is used for illustration)
• A: The matrix $$\mathbf{A}$$ (here a random matrix is used for illustration)

First, the components of $$\mathbf{X}(t)$$ are represented by dummy variables using Unique. Then the function X[t] is created from these. Then the system of equations and the initial condition constraints are stated. Then NDSolve is used to solve the system of equations subject to the constraints.

ClearAll["Global*"];
L = 4;
tEnd = 10;
X0 = RandomReal[{-1, 1}, L];
A = RandomReal[{-1, 1}, {L, L}];
variables = Table[Unique["x"], {L}];
X[t_] = #[t] & /@ variables;
system = X'[t] == A.X[t];
constraints = X == X0;
sol = X[t] /.
NDSolve[{system, constraints}, variables, {t, 0, tEnd}];
Plot[sol, {t, 0, tEnd}] • It's not necessary to define variables. NDSolve can handle implicit matrix. A recent example: mathematica.stackexchange.com/q/230683/1871 Sep 26 '20 at 11:57
• In addition to @xzczd's example, see the docs as well. Sep 26 '20 at 12:14
• Oh yeah. Woops. Thanks!
– Tom
Sep 26 '20 at 14:55