2021 Resolutions

This year I am late to the game but let’s jump right into it.

Health:

• Have my wrist surgery and go through rehab.
• No alcohol except Friday and Saturdays. No beer when possible, stick to raki, whiskey and red wine.
• Maintain: Paleo diet, daily (16 hours) and bi-monthly (2 day) fasts, cold showers, occasional sauna.
• Vit D, K, berberin, resveratrol and NAD booster supplements

Sleep:

This is big. I will try to go to bed around 10-11pm and wake up around 7am. To this end:

• No phone in the bedroom (which forces me to walk to the living room to turn off the alarm)
• and bedtime readings: I will keep these readings light (ethnographies, history, fiction)

If I can accomplish this, this will give me time for a morning routine which is:

• draw for half an hour or walk after my morning coffee
• read on Causal Inference for an hour with my second coffee of the day. And no more coffee after 12pm.

Then my after work routine becomes:

• Code for an hour (maybe for a blog post, or a project or implementing things from morning readings).

Calisthenics:

Since I’ll have a wrist surgery this year and the rehab will take an entire year and even then I will lose mobility in my wrist I have to accept that my calisthenics journey has come to an end. I will try to do legs and core exercises and pick up swimming next year if it is easy on my wrist. The days I exercise abs and legs, I will draw in the mornings, the days I don’t exercise I will go for a walk in the mornings instead.

Blog

I will try to keep the blog more active this year. I want to write two blog posts each month. At least once a month. There already are some topics in the pipeline:

• Finish the basic stats blog posts
• Publish the Naive Bayes classifier for detecting the language of a text.
• Power calculations for one and two sided tests.
• Surrogate metrics and how they can be misleading (with DAGs)
• Instrumental Variables example for non-compliance in tech products.
• Synthetic Control example
• Ratio metrics, delta method and CUPED.
• Causal Identification vs Causal Discovery vs Parameter Estimation.
• Type M error as an illustration of bias/variance trade-off.

Readings

I don’t have a particular list this year.

At night I want to finish some fiction I have been putting off like some Joyce, Moby Dick, The THree Body Problem along with some ethnographies on the Inuit, the Hadza and the !Kung. I also want to read some of my history book on the Dutch Golden Age, Inner Eurasia, Ancient Mediterrenean Trade Routes etc.

For my morning readings I will read Shalizi’s Data Analysis book, re-read Pearl’s Causality Primer and the Causality Mixtape and Imbens and Rubins’ Causal Inference chapters. I will also read some more recent papers that are relevant for my work as a Experimentation Data Scientist.