![]({"https://images-na.ssl-images-amazon.com/images/I/410TgmQuARL._SX398_BO1,204,203,200_.jpg":[400,500],"https://images-na.ssl-images-amazon.com/images/I/410TgmQuARL._SX258_BO1,204,203,200_.jpg":[260,325]})
While generally clear, there there are enough cases where Healy uses terms & notations before explaining them, and at least one where they're not explained _after_ they're used either. If there were a more comprehensive index, or a section explaining these, it wouldn't be a problem, but there isn't.
There's also one case where the text associated with a figure is both inaccurate and essentially "left as an execise for the reader.
The example I'll give is from p.47, but it's not the first instance of this problem -- just the straw that broke my back.
FWIW, I'm not a newbie: I've been a programmer for 40 years, and have worked in at least a dozen high-level languages, including APL, FORTRAN, COBOL, Pascal, C/C++, Java, Lisp, Prolog, and Perl. Some of these (e.g., Perl & APL) have pretty exotic sytaxes, so I'm not unfamiliar with compact and hieroglyphic notation. I'm not bragging -- after 40 years, I'd be a slacker if I hadn't accumulated a decent amount of experience -- just saying that I do have (at least in theory) enough background that I don't think I'm just being stupid.
In this example, Healy has presented a data frame, and then a tibble created from it. He says "Look carefully at the top and bottom of the output to see what additional information the tibble class gives you over and above the data frame version."
The only _substantive_ difference at the top is a row between the first row (var names) and the first observation that reads as follows:
<fct> <fct> <dbl> <dbl>
"dbl" is easy, but WTF is fct? "function"? "fact"? looks like string data (perished, survived; male, female), but my intuition says that we have two defined enumerated data types, each with two values ... but ...
I should not have to rely on guesswork: the meaning of <fct> should be easily accessible.
There's no index entry for "fct"; I see one for "facet" -- is this a facet? I won't find out for another 30 pages. If this were the Kindle version, I could search for <fct>, and for fct if I find nothing for <fct> -- but (a) this if the paperback and (b) I shouldn't _have to_ search for the meaning of a notation in an example.
Returning to the tibble, there is _nothing_ added at the bottom: the last row is the same as the last row of the data frame (modulo the addition of a decimal point after 344, the value of the "n" variable).
I haven't decided yet whether to return it, but I'm put off by sloppy proofreading & copy-editing, partly because it's unprofessional, but largely because I'm a publisher and have _done_ proofing & copy-editing for 6 books, and I am scrupulous about it (and have at least 2 other people check me before I release a title). It's hard, and time-consuming, but to have an otherwise decent book marred by random glitches like this decreases the value & utility of the book.
If I can find errata for this book, and they address the many glitches I've found, I may keep it; o/w I'll wait for the 2nd edition.
- Amazon Business : For business-only pricing, quantity discounts and FREE Shipping. Register a free business account
