Yesterday, I spent some time looking at new keyboard options online. Most of those searches happened via the Best Buy application. My last few keyboards have been ergonomic keyboards built by Microsoft. A replacement keyboard in the same model that I have right now is probably the best option in terms of price. The Natural Ergonomic Keyboard 4000 from Microsoft has been my go to keyboard for years. They now have a couple of different price points and options. One of the options is well over one hundred dollars. Normally, I would be able to head out to a store and take a look at the keyboards. During these strange times in quarantine that is not happening. I’m going to watch a few videos and see what people think about the Microsoft Surface Ergonomic Keyboard. It might very well be my next keyboard. This keyboard I’m using right now might just keep on working and no replacement will be ordered.
Right now my time is being split into two categories of activity. The first category of activity happens to be occurring during my morning window of writing some prose. You normally get to see that in the form of a weblog post. The things that happen during the course of waking up are covered and my ideas are converted into writing about the nature of things. Most of that ends up circling back on the idea of striving toward a perfect possible future and the efforts we make to move forward. The rest of it is muddling prose distilled from inaction and shared with the world. Nobody really wants to read about somebody else being conflicted due to a touch of writer’s block or worse procrastination. The second category of activity happens to be a little bit of Jupyter notebook development. My new Data-Analysis repository on GitHub is devoted to sharing notebooks with simple working examples of things you can do and modify to do other things. The idea is for me to explore data analysis efforts in Jupyter notebooks in a definable and repeatable method.
I’m really focused on putting together a collection of tools for people who are learning to do some complex things in the data science space with machine learning. Getting to that point means helping people get used to opening up rich data sets and doing something with them using Jupyter notebooks. The benefit of this approach is that somebody of any programming skill level can simply pick up the notebook and read the instructions clicking the executable code boxes from the top to the bottom and run the example. That is powerful in nature because reading and clicking is an easy way to start. You can also tinker with each box in the notebook until it does what you want. For somebody learning how to code this really isolates the problem in the chain of commands and you get pretty decent error messaging. Sometimes that is the key to learning enough to overcome the error.