Showing posts from June, 2016

WWDC 16 - Achievement Unlocked

At WWDC 16, I met and discussed Swift 3.0 and Proto Buffers with Chris Lattner, the inventor of the Swift language.

DIY Happiness

I have been working on what very likely is the root of another innovation breakthrough to add to my portfolio. In essence, I believe I have identified a technique on how to mechanically trigger some specific endorphins and dopamines that control habits. That is without any external chemical additive. Aka pills and alike. It has been only one year of research and the science behind it isn’t a piece of cake. However early results don’t need too much interpretation. The subjects' response clearly show that I must have hit at least one important pebble of jackpot road. Anyhow, one important and DOH! Moment led me to realize how much stress and micro-multi-tasking these days has side effects that ultimately generate new symptoms with little science to understand them. This ultimately leads to a pattern of addressing these effects in curing A and instead your issue is B.  What happens is that your big box (brain) adapts to technological stimulus a lot slower than you might think. It re…

It’s a bug of Spatial size!

Spatial correlograms are great to examine patterns of spatial autocorrelation in your data. They show how correlated are pairs of spatial observations when you increase the distance (lag) between them – they are plots of some index of autocorrelation (Moran’s I or Geary’s c) against distance. Although correlograms are not as fundamental as variograms (a keystone concept of geostatistics), they are very useful as an exploratory and descriptive tool.  I use ncf as my default package in R to evaluate patterns where the spacial analysis does what I want for my scenarios. In picking up a good college for my children for example something like the analysis below can give clues in betting on more favorable odds than just pure word of mouth.

With such a basis of knowledge, you can imagine my surprise when this morning I discovered that we have a bug for over 15 years in the statistical analysis we use for brain scans! The disturbing bit is The researchers used published fMRI results, and alo…