

Data Scientist. Drinking iced coffee in Swedish winter.
Many people try to feed data to some fancy models without first understanding their data and company’s business; some people understand their data and company’s business but are not able to communicate results to stakeholders; few people understand their data and company’s business and are still able to communicate results to stakeholders and even wider range of audience.
…a generation process and the question we want to answer requires going beyond just looking at data. This illustrates perhaps the key lesson to learn from Simpson’s Paradox: the data alone are not enough. Data are never purely objective and especially when we only see the final plot, we must consider if we are getting the whole story.
…y to be told that this technology is obsolete and the next thing will make me even more productive. I stopped switching between IDEs (integrated development environments) a while ago and just settled on Jupyter + Sublime Text because I realized the extras were only getting in the way of me writing code.