Wrong summation in code for Figure 2.6

[martinandrovich]

For Figure 2.6, is the summation not wrong here? In the book, equation (2.36) states that:

where in the code (both Python and MATLAB), the value of s is computed as

QuantFinanceBook/PythonCodes/Chapter 02/Fig02_06.py

Lines 31 to 35 in 7b8ca25

	s = 0;
	for i in range(0,len(X)-1):
	s = np.power(X[i+1]-X[i]-mu*dt,2)
	sigma = np.sqrt(s/(m*dt))
	print('Estimated parameters are: mu = {0} and sigma = {1}'.format(mu,sigma))

but shouldn't it instead be:

s = 0
for i in range(0, len(X) - 1):
	s = s + np.power(X[i+1] - X[i] - mu * dt, 2)

In our implementation, we get

$$\hat{\mu} = 0.0014 \quad , \quad \hat{\sigma} = 0.0318$$

using

mu = 1/(m * dt) * (X[-1] - X[0])
s = 1/(m * dt) * np.sum([(X[i + 1] - X[i] - mu * dt) ** 2 for i in range(len(X) - 1)])
sigma = sqrt(s)

[martinandrovich]

Resulting plot:

[LechGrzelak]

hello Martin, yes, you are correct. It is now corrected in both: Python and Matlab. Thanks for pointing that out. Best wishes, Lech

…

On Wed, 15 Jun 2022 at 18:19, Martin Androvich ***@***.***> wrote: For Figure 2.6, is the summation not wrong here? In the book, equation (2.36) states that: [image: image] <https://user-images.githubusercontent.com/4539736/173874916-a2cfaefb-981c-42e9-af16-0ddc33cb17e9.png> where in the code <https://github.com/LechGrzelak/QuantFinanceBook/blob/7b8ca25c837d6040cecd4471d4ea1d1667fd0999/PythonCodes/Chapter%2002/Fig02_06.py#L33=> (both Python and MATLAB), the value of s is computed as s = 0for i in range(0, len(X) - 1): s = np.power(X[i+1] - X[i] - mu * dt, 2) but shouldn't it instead be: s = 0for i in range(0, len(X) - 1): s = s + np.power(X[i+1] - X[i] - mu * dt, 2) In our implementation, we get $$\hat{\mu} = 0.0014 \quad , \quad \hat{\sigma} = 0.0318$$ using mu = 1/(m * dt) * (X[-1] - X[0])s = 1/(m * dt) * np.sum([(X[i + 1] - X[i] - mu * dt) ** 2 for i in range(len(X) - 1)])sigma = sqrt(s) — Reply to this email directly, view it on GitHub <#13>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEJBBMS7MX5ZPCQCCALUUYLVPH7A5ANCNFSM5Y34RU5A> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>