Data Science from Scratch

Excellent book that takes you through a bunch of core Data Science techniques from the ground up using Python giving not only a great overview of many of the mainstays of the field but also offers the user the chance to build an understanding that you just won’t get from plugging data into a pre made library. The python throughout the book is beautifully written and extremely idiomatic and the crash course at the beginning of the book is worth the price of admission alone. My only slight complaint is the black and white printing throughout (unlike some of the other O’Reilly DS books) – colour highlighted syntax would have made the code easier to read.

I had some understanding of concepts and techniques before reading this book. It collects all of these in one place with clear explanations and examples. This book has done a very good job of reinforcing my views and building my confidence. It confirmed my recent decision to stick with Python. (I think Python works best for programmers.) I will now put what I have learned into my mission to finish in the top 1% of Fantasy Premier League.

This is a good book. I didn't know any Python or data science before I started and now I'm about two-thirds through the book and can say that I know some about both...enough to take on various side automation or data analysis projects with Python. The examples are interesting, illustrative and fun to work through. In terms of level of maths, the book is about building your own algorithms rather than using ones in a library, so while you'll be able to get through the book if you are not that good at maths, to get the most out of it you have to be comfortable with probability/statistics and some calculus ideas, or willing to put the time in to get there, but not really anything too intense. In terms of level of programming, I am probably considered a noob by people who consider themselves programmers, but not a noob by people who consider themselves non-computer scientists, and it was perfect for me.

Really helped me get into the practical use of python at work. Having already completed a few courses I needed an easy to read tutorial for a more varied use of python. This book is great and so easy to read I actually read it cover to cover on the train.

I bought this book for my holidays, as I wanted to learn python, and am interested in going into data science after I graduate (I am a mathematics undergrad). The book is good in that it is easy to follow, code-wise, and I have found it sufficient to learn what is going on with python. I would, however, recommend using another resource alongside this book for actually learning to program (if you are new to programming, like me). This is because it's useful to get a lot of practice, and doesn't explain fully how python actually works - this was a problem as I found myself encountering lots of strange bugs which I didn't understand. The mathematical sections are very (very) poor, with no explanations. For me, this was not a problem because I have done most of the stuff before, but for anybody else, this will probably be an issue. My main qualm with the book is that the author tends to define his own functions for lots of trivial situations rather than use modules like numpy and pandas, which is really annoying! The book gives a very small introduction to lots of topics, which is nice, but I would have liked to have seen a bit more content. It is only around 300 pages long, and given the author babbles quite a bit, I feel as though he could've packed more information into what could've been a very useful guide.

Clearly written book in a genre that has plenty of alternatives. * Excellent introduction to data structures throughout * Useful practical considerations of the underlying algorithms * Functional programming style which is unusual in this genre * Lively style by Joel

Excellent insight into data science, a nice primer before one begins to delve into it properly and a good place to start if one wishes to later go into Machine Learning.

This is a great overview book which assumes little. Few things are covered in sufficient depth to make a massive impact, but to give you a general overview of techniques and ideas, it serves the purpose. Good way to learn Python and get some foundations in the field of data science, but only barely skims the surface in my opinion.

Data Science from scratch is must for the beginners who want an overview and theoretical concepts on python, data visualization, data science , ML ,neural networks and so on. It also has a crash course on Python Combining this with Hands on Machine Learning and Tensorflow it is a combination worth spending.

This book is nice to improve the understanding of some details underlying the data science algorithms, but it falls short in the deepness of the content. Some concepts feels rushed and incomplete; the explanation sometimes isn't clear. Even though the book is shallow, I would recommend it; here and there you can get a valuable piece of information from it.

Un libro que merece la pena para todas las personas que quieran adentrarse al mundo de data science, y no tengan un punto de partida, este libro nos ayuda a identificar casos sencillos hasta casos complejos con una ayuda visual del problema.

J'ai choisi ce livre car bien que ma formation universitaire ait contenu 30% de maths, je n'ai pas eu l'occasion de pratiquer beaucoup depuis. Le livre ne nécessite aucun pré-requis en maths à part les opérations basiques, et explique tous les concepts (statistiques, probabilités, algèbre linéaire, etc...) de manière très didactique. Je développe déjà en Python, donc la partie "crash course" m'a surtout servi à comprendre quelle partie de Python est utilisée pour la data science. Je pense que le livre peut etre difficile à appréhender si on n'a pas déjà des bases en programmation. Je recommande ce livre car il explique bien les bases, est plutôt drôle à lire, et contient des pointeurs vers d'autres ressources pour approfondir les sujets.

Un libro muy útil para iniciarse al mundo DS con Python, aunque creo que es recomendable tener un mínimo de conocimientos previos (ya sea en R o haber hecho algún curso online), si no puede ser que no sea tan básico.