I want to solve a PDE as:
$\frac{\partial u}{\partial t}=\frac{\partial u}{\partial x}$, with conditions: $u(x,0)=sin(\pi cos(x))$ and $u(x+2\pi,t)=u(x,t)$. I have solved it by hand and obtained: $u(x,t)=sin(\pi cos(x+t))$. I would like to solve it both numerically and analytically by Mathematica. I have written:
NDSolve[{D[u[t, x], t] == D[u[t, x], x], u[x, 0] == Sin[Pi Cos[x]],
u[x, t] == u[x + 2 Pi, t]}, u, {t, 0, 10}, {x, 0, 5}].
What's the problem?
You had a simple mistake - getting the order of the t and x arguments mixed up in a couple of places, giving you the error
NDSolve::conarg: The arguments should be ordered consistently.
But then you should also replace the equation u[x, t] == u[x + 2 Pi, t] with u[t, 0] == u[t, 2 Pi] and it works just fine
sol = NDSolve[{D[u[t, x], t] == D[u[t, x], x],
u[0, x] == Sin[Pi Cos[x]], u[t, 0] == u[t, 2 Pi]},
u, {t, 0, 10}, {x, 0, 5}][[1]]
And you can compare the numerical and analytic solutions
Grid[{{DensityPlot[u[t, x] /. sol, {t, 0, 10}, {x, 0, 5},
PlotPoints -> 100, ImageSize -> 400],
DensityPlot[Sin[π Cos[x + t]], {t, 0, 10}, {x, 0, 5},
PlotPoints -> 100, ImageSize -> 400]}}]
NDSolve returns a doubly nested list for some reason, {{u -> ...}} ans so when you type u[3,3] /. sol it returns {0.125417}, which seems ugly to me. [[1]]selects the first element of the list, so if I include it int he definition of sol then u[3, 3] /. sol just returns 0.125417. The plotting seems to work just fine without it.
$\endgroup$
DSolve[{D[u[x, t], t] == D[u[x, t], x], u[x, 0] == Sin[\[Pi] Cos[x]]}, u, {x, t}], which gives{{u -> Function[{x, t}, Sin[\[Pi] Cos[t + x]]]}}in version 10.3. $\endgroup$