Computer Science Tips?

Dear Erick,

Here are some things I've picked up along the way:

• One should be able to program very fluently in at least one major programming language. By major, I mean something that is widely used like Java, C++, and in certain fields Python. The analogy here is to a foreign language; you would like to be as fluent as a native speaker. Much like a foreign language, you can get fluent in a programming language by practicing a lot. Write a lot of programs of things you find interesting to do. I wrote small computer games when I was starting out. Then, I got in to entomology and wrote a program to do some taxonomy analysis. It doesn't matter at this point that the code never gets used widely. You just want to gain experience writing code.


• One should learn data structures and algorithms. What is meant here is the study of how information is stored in a computer (data structures) and the study of general concepts of how to make computers do some class of general tasks such as sorting and finding the shortest path between two cities on an idealized map. While you will find most CS majors know the basics, knowing these two well is important. Much like programming languages, you would ideally like to have a good breadth of this at your fingertips. I think it is a really fun field. Machine learning is a branch of algorithms and is a really hot topic now. The canonical book is this one, https://mitpress.mit.edu/books/introduction-algorithms, but please note that it is more for the junior to senior level CS major. I just mention it for future reference.


• Being able to check quickly and well your own code's logic is a critical skill. This is how a coder really stands out. The computer will usually tell you when you make a syntactic error, but will not tell you if you made a logical or typographical error. Checking for logical and typographical error is almost completely in human hands. For example, suppose you meant to add 1 to some number, like y = x + 1, but instead wrote y = x + 2. This sort of mistake is not caught by a computer. Catching your own logic errors is something you gain with experience, as long as you try to understand the root cause when you find bugs in your code. (Believe me, we all code bugs. Catching them is the key.) You can also test parts of your code by feeding the program appropriate inputs values and checking to see if the output matches your expectation. For example, suppose you wrote a program to add 1 to a number. Then you run the program and feed the program the value 3. You expect the number 4 as the output but got 5. Ah hah! You caught a bug. While this is a very simplistic example, the idea of testing the components of a program is vital.


• The second-to-last tip I would give is a little specific to my own experience and may not reflect of, for example, many of my Google colleagues, so please take this with a grain of salt. I found that there is a lot of low-hanging fruit if you know more than one discipline well. For example, you can code and also know molecular biology, you can do bioinformatics or computational biology. (This was my previous job.) Similarly, if you knew engineering, mathematics, and are a good coder, you could check out the Computational Science and Engineering program at UCSB. They have a nice website which tells what they do. Check out the research gallery. (Small world: Prof. Petzold was on my Ph.D. thesis committee.)


• The last tip is make sure to enjoy doing software engineering. Yes, it is a lot of work to learn, but it is also a lot of fun! And, if one is fortunate, it is possible to work on things that millions of people use.

Best wishes!