Aizan reads stuff #2

Sunday March 3, 2019

Okay, it is time for episode #2 of this series.

Writing Docs Is Better Than Having Docs. “Aversion to reading docs isn’t just laziness, it’s pragmatism”. While I agree that it is faster to read Stack Overflow than reading the docs, but Stack Overflow can’t completely replace docs. I also agree here that being good at writing docs makes a coder a better coder. It is like being able to explain it to a yellow rubber duck.

Faster, better, cheaper: the rise of CRISPR in disease detection. At the first glance, I was able to correctly guess what I was about to read further down: the SHERLOCK system with Cas13 protein. Why this method is desirable more than the revered PCR system? SHERLOCK is not sensitive to power outage, hence this tool is better to function in countries like Nigeria. I wrote about SHERLOCK on Caspershire Meta more than a year ago.

Is the Insect Apocalypse Really Upon Us? Short answer: No. It seems like if all animals (human included) are going to vanish, insects are going to vanish the last. They are diversed, they are resilient. If one sub-species within species is going to die, the other will adapt and flourish. Also, the German study that said insects are heading towards extinction, yeah that one was not a systematic study.

Getting Ahead By Being Inefficient. It is not really about being inefficient here. It is more about being able to learn more things in other to adapt to the surrounding, because, the author assumes that total efficiency equals to diminished plasticity to ever-changing environment. In other words, stay hungry, stay foolish.

Open letter to journal editors: dynamite plots must die. Main reason: barplot obscures data points, and in worse case scenario, obscuring the outlier that make the statistics pretty. Alternative? Scatterplot with geom_jitter(), boxplot with geom_boxplot(), ridge plots with geom_density_ridges(), and some other alternatives.

Pioneering ‘live-code’ article allows scientists to play with each other’s results, and the original announcement article on eLife website. At first I thought they were using Beaker, but nope, they use Stencila, which I bumped into not too long ago. Stencila is cool, it can be coupled with other kernels (e.g. R and Python) and charting library (e.g. Plot.ly), and it has its own formula languaged called Mini. Sounds cool, I might try it one day.

Why I Avoid Slack . Good points here, and why I think anything other than Slack is good when the aim is to encourage proper documentations. Not saying Slack is inherently bad, but it has its own use and should not be used as one-shoe-fits-all kind of solution. Maybe this is why Zulip came into life. Also, Slack is a walled-garden, you need to be a member to be able to see what’s inside.

The Spacing Effect: How to Improve Learning and Maximize Retention. Essentially, the article is about Ebbinghaus’ forgetting curve and the concept of spaced repetition to retain memories. In other words, the same material has to be revised 3-4 times in a carefully-spaced time interval to achieve a high retention rate. Flashcard software like Anki might be helpful in this kind of learning adventure.