METROPOLIS ALGORITHM
MCMC sampling a 2D energy landscape
Temperature T
T = 0.40
Step size σ
σ = 0.15
Landscape
Double well
Mexican hat
Rough landscape
Banana (Rosenbrock)
RESET WALKER [R]
CLEAR HISTOGRAM
Accepted
—
Rejected
—
Accept rate
—
Current E
—
Mean E
—
Green dot = current position. Trail shows last 100 steps. Histogram builds the sampled distribution — proportional to e^(−E/T). Higher T explores more; lower T concentrates near minima.
← back to lab index