Table of Contents

Minuit

Tutorials

Google Colaboratory
https://colab.research.google.com/drive/1jufurZXLxzkC1FzyOXpq6I7_-oDQZvCt

Sample

import numpy as np
from iminuit import Minuit
dx = np.arange(10)
dy = ((50.0 * dx) + 100.0) * (1.0 + (0.05 * np.random.randn(len(dx))))
dy_error = np.sqrt(dy)
model = lambda x, a, b : ((a * x) + b)
chisq = lambda a, b : sum(((dy - model(dx, a, b)) ** 2.00) / (dy_error ** 2.00))
fit = Minuit(chisq, a=1.00, b=1.00)
fit.migrad()
fit.hesse()
fit.minos()

fit.values
## <ValueView a=51.49166936856244 b=97.93756640441136>
fit.draw_mncontour('a', 'b', cl=[1.00, 2.00, 3.00])

fit.draw_mnmatrix(cl=[1.00, 2.00, 3.00])

References

https://hackmd.io/@tenoto/Skl_wArqD