BVP Solver Examples¶
Example Code¶
Representative script:
"""IMSL BVPFD example: BVP y'' + y = 0, y(0) = 0, y(pi/2) = 1.
Demonstrates use of bvpfd (finite-difference BVP solver):
y'' + y = 0
y(0) = 0
y(pi/2) = 1
Exact solution: y(x) = sin(x).
Outputs:
- Table printed to stdout
- SVG plot saved to test_output/example_imsl_bvpfd.svg
"""
from __future__ import annotations
from pathlib import Path
from typing import Dict
import matplotlib.pyplot as plt
import numpy as np
from differentialequations import bvpfd
def run_demo_imsl_bvpfd() -> Dict[str, object]:
"""Run IMSL BVPFD example: y'' + y = 0 with Dirichlet BCs.
Args:
None
Returns:
Dict[str, object]: Result dict with keys ``x`` (ndarray),
``y`` (ndarray), ``error_max`` (float), and ``plot_path`` (str).
"""
def ode(x, y):
"""ODE system: [y1' = y2, y2' = -y1].
Args:
x (float | np.ndarray): Independent variable.
y (np.ndarray): State vector of shape (2, N) or (2,).
Returns:
np.ndarray: Derivatives stacked vertically.
"""
return np.vstack([y[1], -y[0]])
def bc(ya, yb):
"""Boundary conditions: y(0)=0, y(pi/2)=1.
Args:
ya (np.ndarray): Solution at left boundary x=0.
yb (np.ndarray): Solution at right boundary x=pi/2.
Returns:
np.ndarray: BC residuals of shape (2,).
"""
return np.array([ya[0], yb[0] - 1.0])
x_mesh = np.linspace(0, np.pi / 2, 30)
y_guess = np.zeros((2, x_mesh.size))
y_guess[0] = np.sin(x_mesh)
result = bvpfd(ode, bc, x_mesh, y_guess, tol=1e-5)
exact = np.sin(result.t)
error_max = float(np.max(np.abs(result.y[0] - exact)))
print("\nIMSL BVPFD Example: y'' + y = 0, y(0)=0, y(π/2)=1")
print("-" * 55)
print(f"{'Parameter':<30} {'Value':>20}")
print("-" * 55)
print(f"{'Domain':<30} {'[0, π/2]':>20}")
print(f"{'Exact solution':<30} {'sin(x)':>20}")
print(f"{'Converged':<30} {str(result.success):>20}")
print(f"{'Mesh points':<30} {result.t.size:>20}")
print(f"{'Max absolute error':<30} {error_max:>20.2e}")
print("-" * 55)
output_dir = Path("test_output")
output_dir.mkdir(parents=True, exist_ok=True)
plot_path = output_dir / "example_imsl_bvpfd.svg"
x_fine = np.linspace(0, np.pi / 2, 200)
fig, axes = plt.subplots(1, 2, figsize=(12, 5))
axes[0].plot(x_fine, np.sin(x_fine), color="#999999", linewidth=2.5,
label="Exact: sin(x)", linestyle="--")
axes[0].plot(result.t, result.y[0], color="#0369a1", linewidth=1.5,
marker="o", markersize=4, label="BVPFD solution")
axes[0].set_xlabel("x")
axes[0].set_ylabel("y(x)")
axes[0].set_title("BVP Solution: y'' + y = 0")
axes[0].legend()
axes[0].grid(True, alpha=0.3)
axes[1].semilogy(result.t, np.abs(result.y[0] - np.sin(result.t)) + 1e-16,
color="#0891b2", linewidth=1.5, marker="s", markersize=4)
axes[1].set_xlabel("x")
axes[1].set_ylabel("|error|")
axes[1].set_title("Absolute Error vs. Exact Solution")
axes[1].grid(True, alpha=0.3)
fig.suptitle("IMSL BVPFD: Two-Point BVP y'' + y = 0", fontsize=13, fontweight="bold")
fig.tight_layout()
fig.savefig(plot_path, format="svg")
plt.close(fig)
return {"x": result.t, "y": result.y[0], "error_max": error_max, "plot_path": str(plot_path)}
if __name__ == "__main__":
run_demo_imsl_bvpfd()
Input (Console)¶
Run the BVP solver script from the package root:
python examples/example_imsl_bvpfd.py
Console Output¶
IMSL BVPFD Example: y'' + y = 0, y(0)=0, y(π/2)=1
-------------------------------------------------------
Parameter Value
-------------------------------------------------------
Domain [0, π/2]
Exact solution sin(x)
Converged True
Mesh points 30
Max absolute error 6.71e-09
-------------------------------------------------------
Plot Output¶
Generated SVG plot:
Description¶
This example solves the two-point boundary value problem:
\[y'' + y = 0, \quad y(0) = 0, \quad y(\pi/2) = 1\]
whose exact solution is \(y(x) = \sin(x)\).
bvpfd (finite-difference BVP solver backed by scipy.integrate.solve_bvp)
is used. The left panel compares the numerical solution against the exact result,
and the right panel shows the absolute error on a log scale.