Regular readers are familiar with the tagline that ends every one of my posts, "As always, I am hochspeyer, blogging data analysis and management so you don't have to." There's a bit of truth there, as well as a bit of attempted humor.
For starters, I've been using "hochspeyer" as my online persona for years. "Hochspeyer" (upper cased H) is a village in Germany where Jennifer and I lived for a few years. My online name (hochspeyer, lower cased h) is something I've been using for years. And although the topic of data doesn't come up as frequently as I'd like, I do try to feature it when possible. One problem I have is that I don't do as much analysis as, say, when I was actually employed as an analyst... for 15 USD/hr.
Pregnant pause for effect.
Yes, I once made only $15/hr working as a data analyst (this might be lots of money in some parts of the world, but in the greater Chicago area, its not- especially when you're feeding, clothing and providing housing for five persons). According to payscale.com, it should have been (at a minimum) 39K per year. According to an inflation calculator, the USD inflation rate between then and now is ~17.6%, so I was roughly 2K under the base salary for an analyst, or roughly 5%. However, at that point in my life, it was the only gig I could score, and the plus side was I learned quite a bit about basic data analysis, Excel, Access, customers, internal and external deadlines, and adding value to both your output and your position.
As a temp, I did analysis for $12/hr. My boss there was a lady who was the Quality Manager. The plant was a contract manufacturer of aluminum die castings for the automotive industry. Although they produced products for all three of the "Big 3" (General Motors, Ford, and Chrysler), their main customer was General Motors (GM). This was when Saturn was an exciting GM product line. The company had made a significant investment in equipment that was specifically designed for small cars, such as the Saturn line. Unfortunately, the Saturn line had disappointing sales for GM, and was ultimately scrapped. I apparently was doing a good job there, however: there were two rounds of layoffs in the company before I was let go. My value add there wasn't special- I integrated Excel data into an existing Access database. I also had the temerity to give reasons when they asked why I couldn't extrapolate results (when one has neither data or experience, one is up the creek without a paddle in terms of validity!)
With overtime, computer crashes and life in general, I haven't had much time to think about data as of late, much less do any analysis. However, the further I get from 10K blog viewers, the stronger the pull is to do a bit of analysis on blog hits to date.
So, here we go: blog speeds and feeds.
I started writing this blog in December of 2012. As of January 2016, I've published 196 posts to this blog- roughly five posts per month. As I'm no professional writer, I've got to say that this paltry volume is a challenge to maintain.
I've been tracking numbers from Day One, and they break down like this:
As far as continents go, only Antarctica has zero readers- Antarctica is 7th.
South America comes in 6th. With the limited amount of data I get back from Google (it's free, so I'm not complaining), I can only guess that language may be a factor in the readership here.
In terms of continents, Australia has the fewest number of countries... 1, and it is predictably 5th in the ranking. However, I think a few more countries could probably be considered a part of Australia for statistical purposes- New Zealand comes to mind immediately, but that's about it. If anyone has any ideas, please comment. The rest of the continents are as follows:
Africa (4th) is actually tied with Australia in terms of total readers, but these readers are distributed among a few countries, so I have to award the spot to Africa.
Europe comes in 3rd, which is slightly surprising because of the widespread use of English there. Twenty-seven countries represent Europe.
Asia is represented by seventeen countries and comes in 2nd, mainly because of Chinese readers.
My top market, again not a surprise, is North America, with a bit over 75% of my global total in readership.
That's it from here for now. Before signing off, I'd like to extend special e Day (2/7) greetings to Prof. Diego Kuonen!
As always, I am hochspeyer, blogging data analysis and management so you don't have to.
Showing posts with label data analyst. Show all posts
Showing posts with label data analyst. Show all posts
Wednesday, February 3, 2016
Monday, September 29, 2014
A cautionary tale of fixed values used in calculating extrapolated fields
Since writing A Tale of Two Databases, I've dusted off my Excel worksheet that I use to keep tabs on blog readership. Yes, deep, deep, way deep down inside of me- somewhere- there's a tiny bit of statistician. And this statistician, who lives somewhere on the Continuum-, er, that is, the Spectrum- likes to keep track of statistics about the blog.
A bit of background is in order at this juncture. I've long been a fan of collecting facts and figures- at least since high school. The year I was a freshman in high school, a new store opened that was on my way to school. The shop was named "The Sutler's Wagon", and Jeff Urquhart (the store has long since closed) was the proprietor. It was a hobby shop. Even back then, I LOVED hobby shops.
Prior to The Sutler's wagon opening, there were three shops I frequented whenever I could get my Dad to take me to them. One was Pilot. I don't remember the whole name of the business, but Pilot was more of a hardware store with a pretty bustling O scale and S scale tinplate selection and HO scale model railroading. Pilot was actually pretty close to Jeff's shop, and was derisively referred to as "the Pirate". Stanton in Chicago was a fairly large store, stocking everything from N to G scale railroading, as well as rockets, RC, plastic models, flying balsa planes... pretty much anything one could ask for in hobbies. The last store was Hill's in Park Ridge, which seemed to have a product range similar to Stanton but smaller selection.
The Sutler's Wagon was different. No trains. A couple of rockets. Some model cars. Everything else was military. Books. Models. Boardgames. Paint. Brushes. Miniatures. Miniatures in 15mm, 25mm, 54mm, 1/35th, 1/72nd, 1/144th, 1/700th, 1/1200th and 1/2400th scale. The 15's, 25's, 1/144th, and 1/2400th were generally diecast metal; the others were generally plastic. He had rules for the games- these miniatures were not just for looking pretty- they were meant for action!
And this has what to do with anything?
I've mentioned this game before. In a very real way, Fletcher Pratt's Naval Wargame taught me long division. While playing the game is fun. I think I had almost as much fun making the ship (data) cards. In the game, each ship has a point value. The point value is based on a number of data which need to be factored into an equation. That's the easy part. From the total point value, one must then create a damage table which shows how a ship's performance is degraded as she takes damage. So, say a ship has a point value of 10,000 and a max speed of 20 knots. The ship takes 1000 points of damage. 10,000/20= 500, so for every 500 points of damage, the ship would lose 1 knot of maximum speed. It's not perfect, and is designed to work best with WWI and WWII vessels, but it does work rather well. There are other factors that can make the game quite involved, but overall it is great fun.
From this point, I started looking at music. On the same block as The Sutler's Wagon was a music store called RPM records. Along with Rolling Stone Records in Chicago, I would scour their bargain "cutout" bins. I discovered a lot of great music in these bargain bins, and by the time I got to college, I had a few hundred albums. My earliest music databases were handwritten on loose leaf paper- I counted an album as being played when I had listened to at least one side.- this worked for both LPs and cassettes. I did have some 8-tracks, but I did not track these.
Flash forward many years. On a normal day I'm on a PC or two at work, and three PCs (simultaneously) at home. In all honesty, a lot of what I do currently is online gaming, but I also write this blog, do a bit of research, maintain a few spreadsheets, and work on my database, Sometimes, I even take a bit of time off to play a game or two on Steam (user= drehenthalerhof).
Here's a lesson for all of you budding data analysts, data scientists, statisticians and anyone else who uses spreadsheets and databases: if the datum doesn't look right, don't assume it is. Check it- even if you are the one that entered it!
So there I was Saturday night, updating the Excel blog readership spreadsheet. This is not a particularly complicated spreadsheet- fifty rows and thirteen columns. As I've repeatedly mentioned, Google's free analytics aren't too deep- they're pretty much stats, and they only show the Top 10. And- because they're real time, if I miss a day or two, I can potentially miss some data. So I've had extrapolate to get a more accurate picture. At some point, though, I accidentally overwrote a formula- and my beautiful table of extrapolated data suddenly was being calculated by a constant instead of a variable! I figured out the error quickly enough, but one can't count on an easy fix every time. When I suspect a problem, or when I need to do a QC on an Excel worksheet, CTRL+~ is my goto.
As always, I am hochspeyer, blogging data analysis and management so you don't have to.
A bit of background is in order at this juncture. I've long been a fan of collecting facts and figures- at least since high school. The year I was a freshman in high school, a new store opened that was on my way to school. The shop was named "The Sutler's Wagon", and Jeff Urquhart (the store has long since closed) was the proprietor. It was a hobby shop. Even back then, I LOVED hobby shops.
Prior to The Sutler's wagon opening, there were three shops I frequented whenever I could get my Dad to take me to them. One was Pilot. I don't remember the whole name of the business, but Pilot was more of a hardware store with a pretty bustling O scale and S scale tinplate selection and HO scale model railroading. Pilot was actually pretty close to Jeff's shop, and was derisively referred to as "the Pirate". Stanton in Chicago was a fairly large store, stocking everything from N to G scale railroading, as well as rockets, RC, plastic models, flying balsa planes... pretty much anything one could ask for in hobbies. The last store was Hill's in Park Ridge, which seemed to have a product range similar to Stanton but smaller selection.
The Sutler's Wagon was different. No trains. A couple of rockets. Some model cars. Everything else was military. Books. Models. Boardgames. Paint. Brushes. Miniatures. Miniatures in 15mm, 25mm, 54mm, 1/35th, 1/72nd, 1/144th, 1/700th, 1/1200th and 1/2400th scale. The 15's, 25's, 1/144th, and 1/2400th were generally diecast metal; the others were generally plastic. He had rules for the games- these miniatures were not just for looking pretty- they were meant for action!
And this has what to do with anything?
I've mentioned this game before. In a very real way, Fletcher Pratt's Naval Wargame taught me long division. While playing the game is fun. I think I had almost as much fun making the ship (data) cards. In the game, each ship has a point value. The point value is based on a number of data which need to be factored into an equation. That's the easy part. From the total point value, one must then create a damage table which shows how a ship's performance is degraded as she takes damage. So, say a ship has a point value of 10,000 and a max speed of 20 knots. The ship takes 1000 points of damage. 10,000/20= 500, so for every 500 points of damage, the ship would lose 1 knot of maximum speed. It's not perfect, and is designed to work best with WWI and WWII vessels, but it does work rather well. There are other factors that can make the game quite involved, but overall it is great fun.
From this point, I started looking at music. On the same block as The Sutler's Wagon was a music store called RPM records. Along with Rolling Stone Records in Chicago, I would scour their bargain "cutout" bins. I discovered a lot of great music in these bargain bins, and by the time I got to college, I had a few hundred albums. My earliest music databases were handwritten on loose leaf paper- I counted an album as being played when I had listened to at least one side.- this worked for both LPs and cassettes. I did have some 8-tracks, but I did not track these.
Flash forward many years. On a normal day I'm on a PC or two at work, and three PCs (simultaneously) at home. In all honesty, a lot of what I do currently is online gaming, but I also write this blog, do a bit of research, maintain a few spreadsheets, and work on my database, Sometimes, I even take a bit of time off to play a game or two on Steam (user= drehenthalerhof).
Here's a lesson for all of you budding data analysts, data scientists, statisticians and anyone else who uses spreadsheets and databases: if the datum doesn't look right, don't assume it is. Check it- even if you are the one that entered it!
So there I was Saturday night, updating the Excel blog readership spreadsheet. This is not a particularly complicated spreadsheet- fifty rows and thirteen columns. As I've repeatedly mentioned, Google's free analytics aren't too deep- they're pretty much stats, and they only show the Top 10. And- because they're real time, if I miss a day or two, I can potentially miss some data. So I've had extrapolate to get a more accurate picture. At some point, though, I accidentally overwrote a formula- and my beautiful table of extrapolated data suddenly was being calculated by a constant instead of a variable! I figured out the error quickly enough, but one can't count on an easy fix every time. When I suspect a problem, or when I need to do a QC on an Excel worksheet, CTRL+~ is my goto.
As always, I am hochspeyer, blogging data analysis and management so you don't have to.
Subscribe to:
Posts (Atom)