Understand — Learn what the new product/feature is all about, its functionality, user experience, objective, timeline. Brainstorm as required to vet out the details. It helps to come up with data requirements, precision/recall expectation, success metrics & counter metrics.
Evaluate — Perform quantitative studies to size the impact, study feasibility…
Here are some key skills you should hone to become a good analyst:
Approximation thru Triangulation: You remember (or have it handy) the key business metrics and are able to leverage that to guesstimate the impact for any product/feature or bugs.
Hypothesis Formation: You closely follow day-to-day business though networking and subscribing to various reports & internal communication. You can leverage that to come up with hypothesis when your metrics move.
Validation of Hypothesis: This is where your technical skills get used. Use SQL or Python to dig into data to prove/dis-prove hypothesis. Learnings go into your knowledge base for future triangulation.
Story Telling : This is where your presentation skills get used. Convert your data and findings into clear charts & insights and communicate to leadership. In addition to sharing insights from past data, highlight opportunity. This is what gets leadership team excited the most.
Below are some tips for you to improve/optimize performance of your Pandas operations. These tips are helpful when you have lot of data or lot of processing or both. I have linked to some source articles where you can learn more about each topic.
GIT is a version controlling system used in software engineering. For Data Scientists working in teams, I recommend using it for especially for your python code. Though you can check-in Jupiter notebooks, GIT is not very helpful for collaborating with Jupyter notebooks — in fact, I am not aware of any better collaborative environments for Jupyter notebooks. Please leave a comment if something works for you.
With GIT, it is easy to find common commands to perform certain actions. But when mistakes happen (mistakes do happen), it is hard to find how to undo what you just did. This is page serves as a reference on how to undo common tasks on GIT.
This is a very practical course on A/B testing by Udacity & Google and I have recommended this to many folks interested in A/B testing. Here are my original notes taken from this course. Course has 5 sections: