Ok, a bit about my background in cliffs form.
BS in Business Management, Economics/Econometrics field
MS (all but thesis right now) in Economics with PhD core (killed math and micro prelim, failed macro prelim)
Thesis is an optimal control problem with hazard functions, jointly solves for an action and a time, bellman equations, all that, written in Mathematica. Should be done soon.
Worked as a research assistant putting together fiscal impact models (OLS, Time Series, 2SLS, 3SLS) and doing basic testing on them (things like granger causality, looking for heteroskedasticity, stationarity, all that but I mean I don't remember a lot of it. Used STATA so it was fairly...canned)
After/alongside my MS coursework I took a lot of mathematics, like 3 Real Analysis courses, a senior level Matrix Theory class, Abstract Algebra, Proofs (obviously for RA), ODE to see what was after calculus and did very, very well at that stuff, so I'm sound at abstraction.
Been working in IT for a year and 10 months or so. TL;DR my wife was diagnosed with an autoimmune disease and living as poor grad students was something I couldn't do to her any longer, so I took a good job with health insurance.
I do information security engineering for regulatory compliance controls on user accounts and computing environment. So it's half soft/business-y and half technical.* While doing this I was promoted early, learned SQL, Python, and R (not amazing at any of them but comfortable and love them all) to get some stuff done. SQL daily, Python weekly, R sparingly when people want me to look through their data or management wants metrics. Nothing killer, mostly GGPLOT and ODBC/SQL. I've played with Pandas, Numpy, and Matplotlib in Python but most of it is for automating regulatory compliance issues and reading/writing to SQL tables.
Currently taking Machine Learning on Coursera, planning to do the Specialization when it starts.
This week, one of the programmers on the team left, and I've been offered a spot on the development team since they know I like it and they can teach me to do it for 'real'. I've accepted since a) I love to learn and b) I think my quantitative and modeling background is solid, but my programming could use some improvement if I want to get into data science. It won't be scientific programming, but I'll become solid at Python, JS, SQL/NoSQL and working with different data structures like JSON and ugly flat files as well as working primarily in Linux.
Am I doing this right?
Aside: I'm worried that by the time I'm 'done' with learning programming, the fad will have come to a head and the industry saturated and extremely competitive. I feel like I'm just OK at a lot of this stuff, but I love it all. That said, I don't want to be a pure programmer for life and will eventually of course need the quantitative component, since math/modeling/explaining things with math and modeling are my true interests. In a way, piping economics skills to things that aren't stereotypical like trading stocks.
[link][18 comments]