A voter fraud theory strikes again! I have come to rather enjoy these, as if you just take a deep breath and remember that there was no fraud, these amount to an interesting puzzle to solve. It's just an episode of Penn & Teller "Fool Us" where the goal is to figure out how the magician does his trick. Instead of magic, we have math and data as the methods at work.

You can find the theory of Dr. Douglas G. Frank presented in a recent legal brief from Attorney Matthew DePerno (attorney for the Plaintiff, William Bailey v. Antrim County). You can also listen to him walk through the brief in an interview on rumble.

Read more…

I've been presented with another theory of fraud in the 2020 Presidential election. Have a look at the video, which compares snapshots of the election results as extracted from NY Times .json files to "prove" that votes were switched and lost. This is done via some python code we'll take a look at below. The video walks through this Natural News post in a very lively and shout-splaining fashion. Bold claims are made, including:

• NY Times .json files are analogous to a "Flight Data Recorder (black box)" when it comes to being official transaction logs of election night results
• over 500k votes were switched from Trump to Biden
• Bannon and Giuliani have confirmed whistleblowers from Dominion, a ballot counting software company (that was on 11/11; the pattern is always the same: dangle some mind-blowing tidbit and then we never hear about it again? Nevertheless, for weeks I'll hear people citing "those Dominion whistleblowers they found")
• a fraud simulator (spoiler: it just switches key states when you toggle between fraud and no fraud scenarios, with no explanation of how it's calculating anything at all)
• Trump and DoD personnel watched this fraud happen in real time, have it recorded, and allowed the fraud to happen so they'd have proof (but still filed lawsuits in five states, just to withdraw from two)

In any case, we'll be focused on the data analysis and short answer: no, there's nothing here folks.

Read more…

While the 2020 election is behind us, this is not the case with interpretations. Theories abound proposing how and why we can know that fraud is the overwhelming conclusion. One such theory is contained in this video by a Dr. Shiva. The gist goes like this:

• in certain states, one may vote "straight ticket," which means making one choice to select all same-party candidates on the ballot
• one may also "direct vote," meaning one makes individual choices (which may or may not be for all same-party candidates)
• for a given party, there is an expected correlation between the percent of straight ticket votes in an area (e.g. precinct, county) and direct votes for others of the same party; deviating from this correlation implies fraud

Let's take a look!

Read more…

I've used Arch linux for about 8 years from what I can tell. I love the minimalistic nature, and also think pacman and pacaur are pretty great for package management. Arch doesn't give me stuff I don't need, stays out of my way, and has fantastic documentation. That said, at work I need Ubuntu due to some pre-built binary availability and the lack of desire to build all of them each time they update. I decided to go with both so I could use Ubuntu for specific tasks and Arch for my daily driver.

I was curious if I could use btrfs to create one main shared space for both, choosing which subvolume to mount at boot. This eliminates the need to partition the disk and guess sizes for each. I also encrypt my drives, so it required the ability to create an encrypted container on btrfs and point the bootloader to that specific container based no which distro I was booting. Lastly, my newest company issued computer shipped with UEFI and Secure Boot, which was entirely new to me.

About ~15hrs later, I've got base systems installed and am feeling pretty happy.

Read more…

As mentioned in a previous post, I've been trying to pick up python, coming from mostly using R over the past 5-7 years for data analysis, stats, and a lot of plotting. I found python's plotting landscape quite a bit more confusing than I expected, with tons of options compared to the typical reigning R champions: base, lattice, and ggplot2.

Granted, there are a ton of other R contenders as well, such as plotly and rCharts, but the three above are the most common. I stick almost exclusively to ggplot2, and rarely find something I can't do (granted, this is typical science-y stuff, not infographics or really complicated stuff). To educate myself, I went through and plotted a pretty standard dataset in various ways to see how one would do it in a bunch of python libraries:

• ggpy
• matplotlib
• seaborn
• plotly
• cufflinks
• altair
• bokeh
• pygal

Read more…

I've been using R for at least 5 years. Lately, I've been trying to pick up python as well, which is quite the learning experience! I've been through some tutorials (or pieces of them) from think python and lpthw, as well as classes on Coursera. With some basics under my belt, I started getting into numpy and pandas, which is where things got tricky. I can vaguely recall a time in my R Padawan days when I had a sense of what I wanted to do, but didn't know enough to even formulate the question correctly. I also find I don't learn well by just reading docs. I want to poke and prod around, trying to wrap my head around what I'm reading. For example, I ran across the the function tolist() via SO, but have no idea how to interpret this as it is:

Series.tolist()
  Convert Series to a nested list

I thought a pd.Series was just like an R vector; why would it be "nestable" in the first place? What is a nested list and why do I want that? Some examples just show that it converts from a pd.Series to a list. I don't see any "nesting," but then don't know what that is.

In any case, the point is that instead of trying to just learn pandas in some linear way, I thought it might be more effective to simply take some R code and translate it to python. I know what I want to know and the tools I would use in R, so getting my brain around how to do the same in python seemed like an interesting way to learn. Whatever I pickup will be relevant (I'm already doing these sorts of tasks), and the learning is more targetted (I'm extracting the small % of information I need vs. trying to digest things I likely will never use).

Read more…