Author: Paul van der Laken

What does a tech lead do? – by Jake Voytko

What does a tech lead do? – by Jake Voytko

According to Jake Voytko, data science and engineering teams run more efficiently and spread knowledge more quickly when there is a single person setting the technical direction of a team. The so-called tech lead.

Sometimes tech lead is an official title, referring to the position between an engineering manager and the engineering team. Oftentimes it is just a unofficial role one grows in to.

Now, according to Jake, you can learn to become a tech lead. And you can be good at it too. Somebody has to do it, so it might as well be you! It could allow you to leverage your time to move the organization forward, and enables you to influence data science or engineering throughout the entire team!

In this original blog, which I thoroughly enjoyed reading, Jake explains in more detail what it takes to be(come) a good tech lead. Here just the headers copied, but if you’re interested, take a look at the full article:

  • Less time writing code
  • Helping others often (esp. juniors)
  • Helping others first
  • Doing unsexy, unthankful work to enable the team
  • Being an ally (of underrepresented groups)
  • Spreading knowledge, or making sure it spreads

And this is what Jake feels his work week looks like as a tech lead:

Snapshot from the original article

Cover image via

100 Python pandas tips and tricks

100 Python pandas tips and tricks

Working with Python’s pandas library often?

This resource will be worth its length in gold!

Kevin Markham shares his tips and tricks for the most common data handling tasks on twitter. He compiled the top 100 in this one amazing overview page. Find the hyperlinks to specific sections below!

Quicklinks to categories

Kevin even made a video demonstrating his 25 most useful tricks:

Building a new desktop!

Building a new desktop!

I recently decided to buy a new computer.

While looking for laptops, it struck me that they can be so expensive for the hardware you get. I actually don’t need to my computer to be mobile, as most of the time it just sits in my study.

Hence, I opted for buying a desktop. And even better, I decided to build one myself!

I thought building a PC was going to be all complex and technical, but it’s actually really easy! I hope I can inspire you to try out for yourself as well.

Basically, you need only need 6 parts to build a computer:

  1. Casing
  2. Power supply
  3. Motherboard
  4. Processor (CPU)
  5. Hard drive (SSD)
  6. Memory (RAM)
  7. Optional: Graphics card (GPU)
  8. Optional: (extra) Fans
Desktop Computer Components (With images) | Computer history, Old ...
Via Pinterest (look at that old school case & speakers)

So I did some research into what hardware to buy. Specifically, I wanted a PC that could handle some deep learning and some of the newer video games. Hence, I decided on this setup:

  1. Casing: Be Quiet! Base with pre-installed fans
  2. Power supply: Cooler Master V550 Gold
  3. Motherboard: MSI B450-A Pro Max
  4. Processor (CPU): AMD Ryzen 5 3600X
  5. Hard drive (SSD): Crucial P1 1TB
  6. Memory (RAM): Crucial Ballistix 3200MHz 2x8GB (I got grey ones)
  7. Graphics card (GPU): MSI GeForce RTX 2060 Super Armor OC

Note: these are affiliate links.
If you buy a similar setup, it will generate a few bucks used to keep my website live!

My new setup put together

My setup totalled to about €1100 or $1200, but it may depend on the vendors you pick. Nonetheless, the CPU and the GPU are definitely the most expensive (and important).

I did not buy any additional fans, as the Be Quiet base already had some pre-installed. However, I think it might be better to install extra’s.

Actually, it’s very easy to upgrade (or downgrade) your system. You can easily switch out modules to decrease or increase the performance (and cost). For instance, you can install another two memory cards on your motherboard, or simply spend more on a GPU.

After everything was delivered to my house, I thought the hard part started: building the desktop and putting everything together. But actually, this only took me about an hour or two, with the help of some great tutorials on Youtube:

I hope this convinces and helps you to build your own system at home!

David Robinson’s R Programming Screencasts

David Robinson’s R Programming Screencasts

David Robinson (aka drob) is one of the best known R programmers.

Since a couple of years David has been sharing his knowledge through streaming screencasts of him programming. It’s basically part of R’s #tidytuesday movement.

Alex Cookson decided to do us all a favor and annotate all these screencasts into a nice overview.

Here you can search for video material of David using a specific function or method. There are already over a thousand linked fragments!

Very useful if you want to learn how to visualize data using ggplot2 or plotly, how to work with factors in forcats, or how to tidy data using tidyr and dplyr.

For instance, you could search for specific R functions and packages you want to learn about:

Thanks David for sharing your knowledge, and thanks Alex for maintaining this overview!

Visualizing and interpreting Cohen’s d effect sizes

Visualizing and interpreting Cohen’s d effect sizes

Cohen’s d (wiki) is a statistic used to indicate the standardised difference between two means. Resarchers often use it to compare the averages between groups, for instance to determine that there are higher outcomes values in a experimental group than in a control group.

Researchers often use general guidelines to determine the size of an effect. Looking at Cohen’s d, psychologists often consider effects to be small when Cohen’s d is between 0.2 or 0.3, medium effects (whatever that may mean) are assumed for values around 0.5, and values of Cohen’s d larger than 0.8 would depict large effects (e.g., University of Bath).

The two groups’ distributions belonging to small, medium, and large effects visualized

Kristoffer Magnusson hosts this Cohen’s d effect size comparison tool on his website the R Psychologist, but recently updated the visualization and its interactivity. And the tool looks better than ever:

Moreover, Kristoffer adds some nice explanatons of the numbers and their interpretation in real life situations:

If you find the tool useful, please consider buying Kristoffer a coffee or buying one of his beautiful posters, like the one above, or below:

Frequentisme betekenis testen poster horizontaal image 0

By the way, Kristoffer hosts many other interesting visualization tools (most made with JavaScript’s D3 library) on statistics and statistical phenomena on his website, have a look!

Tutorial: Demystifying Deep Learning for Data Scientists

Tutorial: Demystifying Deep Learning for Data Scientists

In this great tutorial for PyCon 2020, Eric Ma proposes a very simple framework for machine learning, consisting of only three elements:

  1. Model
  2. Loss function
  3. Optimizer

By adjusting the three elements in this simple framework, you can build any type of machine learning program.

In the tutorial, Eric shows you how to implement this same framework in Python (using jax) and implement linear regression, logistic regression, and artificial neural networks all in the same way (using gradient descent).

I can’t even begin to explain it as well as Eric does himself, so I highly recommend you watch and code along with the Youtube tutorial (~1 hour):

If you want to code along, here’s the github repository:

Have you ever wondered what goes on behind the scenes of a deep learning framework? Or what is going on behind that pre-trained model that you took from Kaggle? Then this tutorial is for you! In this tutorial, we will demystify the internals of deep learning frameworks – in the process equipping us with foundational knowledge that lets us understand what is going on when we train and fit a deep learning model. By learning the foundations without a deep learning framework as a pedagogical crutch, you will walk away with foundational knowledge that will give you the confidence to implement any model you want in any framework you choose.