Two-dimensional harmonic oscillator
This tutorial solves a two-dimensional problem with the input-file driver: the isotropic harmonic oscillator, whose levels are degenerate. In atomic units,
\[\hat H = -\frac{1}{2}\left(\frac{\partial^2}{\partial x^2}+\frac{\partial^2}{\partial y^2}\right) + \frac{1}{2}\,\omega^2\left[(x-x_0)^2+(y-y_0)^2\right], \qquad \varepsilon = \omega\,(n+1),\quad n = n_x+n_y ,\]
and the level $n$ is $(n+1)$-fold degenerate. The same file describes two one-dimensional particles of equal mass in a common harmonic trap.
All files are in src/isotropic_2d_harmonic_oscillator_potential/ of the FEMTISE_TUTORIAL repository.
1. The potential
function qho_2d(x, params::Tuple)
ω, x₀, y₀ = params
return 0.5 * ω^2 * ((x[1] - x₀)^2 + (x[2] - y₀)^2)
end2. The input file
The 10 lowest levels with $\omega = 1$ on the square $(-5, 5)^2$, divided into $100\times100$ cells of quadratic elements:
full_path_name = ./output/qho2d
dom_type = s
nev = 10
dimension = 2D
sigma = 0.0
adhoc_file_name = adhoc_potential_function
potential_function_name = qho_2d
params_potential_types = f f f
params_potential = 1.0 0.0 0.0
analysis_param = false
output_format_type = jld2 eigen
## ONLY FOR 1D EIGENPROBLEMS
L =
Δx =
## ONLY FOR 2D EIGENPROBLEMS
Lx = 10
Ly = 10
nx = 100
ny = 100
different_masses = false
reduced_density = false3. Running the tutorial
julia --project=. setup.jl # once
julia --project=. src/isotropic_2d_harmonic_oscillator_potential/run_tutorial.jlThe script solves the problem (unless results already exist in ./output), compares the eigenvalues with the exact degenerate spectrum and samples the densities on a regular grid:
id, results = simulate("input")
ε, order = sorted_spectrum(results)
ϕ = results.ϕ[order]
exact = sort([ω * (nx + ny + 1) for nx in 0:10 for ny in 0:10])[1:length(ε)]
save_figure(plot_spectrum(ε; exact = exact, index_offset = 1,
index_name = "State index", index_symbol = "k"),
"figures", "qho2d_spectrum")
grid = collect(range(-4.0, 4.0, length = 161))
densities = [abs2.(sample_on_grid(ϕ[k], grid, grid)) for k in 1:6]
save_figure(plot_density_maps(grid, grid, densities, titles; columns = 3),
"figures", "qho2d_densities")4. Results

Spectrum. (a) The computed eigenvalues reproduce the levels $1, 2, 3, 4$ with degeneracies $1, 2, 3, 4$. (b) Relative error, below $5\times10^{-6}$ for all ten states. All eigenvalues are upper bounds.

Eigenfunctions. Densities of the six lowest states. Within a degenerate level any linear combination of eigenvectors is also an eigenvector, so the orientation of the nodal lines of states (b)–(c) and (d)–(f) is arbitrary: the solver returns some basis of each degenerate subspace. Quantities that are invariant within the subspace, such as the total density of a level, do not depend on this choice.
5. Going further
- Different masses. Setting
different_masses = 0.25gives the second coordinate a mass four times larger ($p_2 = 0.25\,p_1$). The levels become $(n_x+\tfrac12)\,\omega + (n_y+\tfrac12)\,\omega/2$ and the degeneracies change. - Reduced densities. With
reduced_density = truethe driver also computes $\rho^{(1)}(x)=\int\rho\,dy$ and $\rho^{(2)}(y)=\int\rho\,dx$. - Orthogonal basis of a degenerate level. The vectors returned for a degenerate level are linearly independent but not necessarily orthogonal; see the post-processing guide.