These are the reference I would recommend to people getting into the field. They are the books and online courses that I have found to be most useful in the long run.
These books are all tool-specific. I use Python and Web technologies to build prototypes. This has been extremely useful because I had to build up a data science team and a big part of the job was to get people excited about the potential.
- Hands-On Machine Learning with Scikit-Learn and TensorFlow. This is the best bok I have read on the Python stack for machine learning. It strikes a good balance between theory, practice and code example.
- Effective Computation in Physics: Field Guide to Research with Python. This books cover most of the basics. It teaches you have to use the terminal, Make, and a all the basic python packages for scientific computation in Python. I reccomend is a good allround introduction to the Python stack for Data Science.
- D3.js in Action and Interactive Data Visualization for the Web. It is possible to get people excite about tables and static graphs if you do the presentation right – but it is so much easier if you are demoing a fully interactive product. You can show how your results could be used.
- Flask Web Development: Developing Web Applications with Python. This book is useful because it allows you to build a functioning prototype of a website.
- Clean Code, Code Complete or The Pragmatic Programmer. The general idea here is just to read a book about how to write maintainable code. This is especially important because a lot of data scientists have a background in the natural sciences where the approach to coding is: when it seems to work it is done.
I plan to keep adding to this list as more useful reference emerge.