I liked this series and this part. I think it's important for people using data science in the industry to continue giving insight into best practices, feedback to academic programs, and occasional insights into the problem applications. In my mind, this ultimately improves the quality, education, and marketability of data science.
I discovered the series earlier today on HN and the discovery could not have been any timely-er. I am just about to embark on the first six to eight months of a learning journey and see immediate utility in insightful series such as this one.
I also came across a really helpful post that gives recommendations on progress markers for the self-taught developer [A Better Way to Learn Programming? Notes on The Odin Project;http://everydayutilitarian.com/essays/notes-on-the-odin-proj...]. Guides like these, while they take a lot of time to write and refine, are complete lifesavers for entry-level professionals and prospective practitioners (and especially if they come from professionals that have been "tried and tested").
You are a data wrangler? Perhaps a guide would do; noobs like me have no perspective on what to learn and how to learn it. I mean, it wasn't even a year ago that I was convinced that I could go from 0 to 100 data science-wise in under a year: I wanted to learn it all. It took me the better part of a year to realize that I had wasted innumerable hours devising a curriculum and timelines that were plain dumb. A practical guide could have spared me a lot of hurt and while I cannot at this moment compensate you (or the community for that matter), I am sure that opportunities will certainly arise for me to pay my debt. Would love to see a guide from you- it would come with the added advantage that you would be accessible to the brilliant HN community. I would give away a limb to see such a discussion go down: what to learn and where to learn it from (a lot of folks, I imagine, would not mind recommendations for openly accessible material; I know I wouldn't mind that)? how fast should you expect to go/move/learn? time commitments? tools and frameworks? motivation hacks? where would I go to find remote jobs? what level of proficiency should I achieve in the first sprint?
The best way to learn how to wrangle data is practice, especially outside of academic settings, where the example data is not necessarily reflective of real-world data.
Helping others wrangle data is one of the reasons I publish my Jupyter notebooks open-sourced. A few examples my data wrangling with R:
The "Analysis of Lead Gender and Box Office" project put a smile in my face because it looks like so much fun- that is something that I would like to know how to do in the coming months. Thank you for open sourcing the notebooks and the recommendation to practice then demo on real data.
