This book is not terrible but it could have been a lot better. The English is poor (the author is Italian, I think - certainly not a native English speaker, anyway) which makes it harder to follow what's going on - though not impossible. But here's the real problem; I am still spending more time looking things up online than in either of the two books on Python I have. This, to me, means they can't be good. By contrast, Mastering Matlab 7 by Duane C. Hanselman and Bruce Littlefield, is excellent and my copy is showing signs of imminent collapse through heavy use.
What's the difference? Total information content and it's presentation. The Matlab book contains vastly more info and much of it is concentrated in tables. My copy has a profusion of little post-it notes marking pages with reference tables that I use frequently. These are conspicuous by their absence in Python Data Analytics, except in the case of a comprehensive appendix on Latex codes for use in plot annotation.
My search for a decent Python/SciKit reference goes on.