code review in data science

The original English address is:  https://towardsdatascience.com/15-topics-to-consider-as-you-review-code-in-data-science-10eff0182e71

Here we only record the aspects that need to be considered in code review in data science. It is recommended to read the original text.

Exposure to New Ideas and Technologies

1. dispel doubts
   1. During code review, if you don't understand something and why, you can consult the developer. Understanding removes assumptions and doubts from data.
2. Find ways to clean up and minify your code
   1. If everyone's processing methods are similar, we can discuss whether we can develop a more concise and general tool.
3. discuss next steps
   1. New ideas and train of thought may emerge after the discussion. 1+1>2
4. Explain new libraries or technologies used in the code
   1. broaden horizons
5. understand different ways of doing the same thing
   1. When conducting a code review, you can think about what you would do, why did he do it? Is there a better way.

Feedback on Code Development and Progression

1. Discuss the structure of the code
2. Check for loops and nested loops in your code
   1. Could be more readable and reusable
3. View kwargs and variables
4. Evaluate code output to reduce errors
5. Find logic errors and bugs

Forum to Discuss (in the form of code review, a meeting of two or three people is recommended)

1. Open your heart and discuss what you don't understand and what you don't understand.
2. Seek advice, and ask people if you're not sure which course of action is better.
3. In addition to discussing code, discuss next steps.
4. Share bugs, if you spend 2 hours on a problem and you can't solve it, you can ask for help.
5. Discuss everyone's ideas on the whiteboard, working through some simple examples.