astroml / text_errata

A Catalog of errors and typos in the book "Statistics, Data Mining, and Machine Learning in Astronomy

Text Errata

A List of errors and typos in the second edition of the book Statistics, Data Mining, and Machine Learning in Astronomy

For the first edition, please go to the old errata page

There are a couple ways to submit a new error report

• If you are a GitHub user, you can open a Pull Request which edits this document.
• If you are not a GitHub user, please email the authors directly with your reported error, and we will add it to the list.

Errata

Errors are listed by chapter, followed by page number.

Chapter 8

Page 317: The text after Eq. 8.19 should read in two instances (Y - M \theta) instead of (Y - \theta M)

text_errata's Issues

Figure 3.3 might be fine as printed, despite erratum

Am I wrong? See my comment on the original commit to the README.md file.

Eq 8.11

Shouldn't eqaution 8.11 for the $\sigma^2$ be divided by $(N-2)$ ?

Book erratum: Different distances

I am preparing a class where I am going to use the Statistics Data Mining, and Machine Learning in Astronomy book. I think I run into something strange in section 2.5.2. I hope this is the right place to address this.

The distances calculated by the easy_nn and vectorized_nnare not the same. I think it lacks a square root at equations 2.4, 2.5 and 2.6.

I set up a notebook showing this.

Eq. 9.21 incorrect?

Not positive, but I think the argument of the argmax should read:

\ln \pi_k - \frac{1}{2}\sum \left( \ln(2\pi (\sigma_k^i)^2) + \frac{(x^i - \mu_k^i)^2}{(\sigma_k^i)^2} \right)

since it is just the logarithm of a Gaussian?

Dubious units in Equation 8.77 and 8.78?

I'm going out on a limb here, but...

Equation 8.77 and 8.78 seem to have strange units- y_units^(0.5).
Should the argument of the summation over i be squared?
I am using the digital online edition of the book accessed through EBL at my library, and have not checked the paper copy of the book.

Typo on Page 427

On page 427, in the first paragraph of section 10.3.1, it says omega = 2 pi f = 2 pi P. Should be 2 pi / P.

Eqns 8.67 and 8.68

Hi,

I wonder if eqns 8.67 and 8.68 contain a similar error. Should sigma_i^2 not appear in the second term of the likelihood function? The second term reflects the background component, so why should the statistics of it have anything to do with sigma_i?

Cheers,
Joel Roediger

Chapter 7 (v2), Page 296, Figure 7.6

Page 296 in Chapter 7 refers to Figure 7.6 (which got moved from v1), but this is actually Figure 7.7 and it should be the right panel, not the middle panel. But really, the figure should be changed such that NMF is the middle and ICA is the right. Note that the text was apparently wrong in v1 as well.