How I Model It

Tips, tricks, and insights for working with Python, R, and all things data-ish.

Hello World!


To those of you who've found your way here,

My name is Andrew Bibian and I'd like to welcome you to my site. While it might not look very pretty right now or even have more than one page I'll tell you that that's the point (keep reading to learn more).

I've just started this site using my new server and the one click Django installation from Digital Ocean. The goal is that this blog will be documenting the evolution of my site along with post of other bits of information that I find interesting.

As an introduction to me, my original life plan was to become an ecologist but along the way I realized I just like creating things from code. Training wise I'm a quantitative ecologist with a knack for statistical modeling, scientific computing, and stochastic simulations. Most of my experience stems from writing statistical models that take a Bayesian approach to parameter estimation (think rStan, JAGS, OpenBUGS, WinBUGS). And even though I'm pro Bayes I can/will implement similar models in a Frequentist framework (i.e. likelihood estimation only...no priors).

Even though I'm still very much excited about statistics and modeling (hence the name of this site) in general I'm starting down a new path professionally. In five years the current plan is to become a data scientist for a top company and hopefully in ten years a chief data scientist. But now that I've completed grad school I've quickly learned that being a good data scientist requires a huge knowledge base that includes traditional statistics, machine learning, database design & use, domain knowledge, programming/coding/scripting, version control, automation, dashboard creation, interactive data viz... and the list goes on. And since my goal is to become a data scientist this blog will be dedicated to my learning journey about all things data and tech related.

As of right now, If your curious, I'm working on project with a small team to create a centralized database that brings together hundreds of individual datasets related to a sub-discipline of ecology (https://github.com/bibsian/database-development). While this sounds simple enough it has actually been one of the biggest learning opportunities I've ever had in a job. I've had to learn Python to design software for cleaning, formatting, and uploading data to our database (PostgreSQL), including the slew of task involved with this i.e. unit test, design patterns, remote servers, automation, bash, etc. In the end this database will be open to the scientific community and we'll even make it easier for people to access with an open source R package that's currently in development. Look for the release of out package and database in February of 2017.

If you'd like learn along with me then come back soon for new post or feel free to drop a line at ajbibian@gmail.com (sorry...I don't have a 'contact' page up yet :-/).

Cheers,
Andrew