[Interview]: Dason - that guy that isn't a bot

Dason

Ambassador to the humans
#21
Are you from Iowa, or do you just work there, and you are actually from outer space?
I'm originally from Wisconsin. I did my undergrad in Minnesota. And I'm in Iowa because Iowa State has a really good Statistics department. So I guess I'm in Iowa just for the school. We probably won't stay in Iowa after I graduate but it's possible. Although with the way things are going I'll most likely continue my tour of the Midwest and head over east into Illinois (Noooo!!!)
 

noetsi

No cake for spunky
#22
I'm assuming you're talking about my hatred of raptors. That is simple - they are pure evil. They are killing machines that will stop at nothing to destroy humanity. They also tell terrible gossip.

To be honest it's a mixture of being freaked out in the scene from Jurassic Park where the raptor jumps through the ventilation ducts unexpectedly and my former main online community being the xkcd forums. In case you didn't know: xkcd is notorious for raptor paranoia.
There haven't been living raptors for tens of millions of years....
 

Dason

Ambassador to the humans
#23
That's exactly what the raptors want you to think.

Just so it's clear - I don't actually think there are raptors around. But it's fun to pretend... especially when trinker goes along with it and sides with the evil enemy... **** raptors always being jerks like that...
 

noetsi

No cake for spunky
#24
The evolved into those who wrote comprehensive exams (with the most powerful becoming those who wrote prelims). Note the special glee they get from making these closed book, when in reality no one would ever answer questions like this from memory.
 

Dason

Ambassador to the humans
#25
The evolved into those who wrote comprehensive exams (with the most powerful becoming those who wrote prelims). Note the special glee they get from making these closed book, when in reality no one would ever answer questions like this from memory.
Depends on the question ;)

I know when I was doing consulting I would look at a lot of output that people would bring to the meeting and I would do a bunch of stuff off the top of my head to check if their output even made sense.

Now doing the theory off the top of my head... I typically don't do that but it's still good to know. And Lord knows I wouldn't have taken the time to memorize some of those theorems as thoroughly as I did if the prelim would have been open book.
 

noetsi

No cake for spunky
#26
I dont think I would have the courage to run any complex method without looking something up. Like say that the interaction and error term in a random effect ANOVA are confounded so you can't calculate interaction terms for random effects. Or at least that is what my class slides say. When I did HLM which also uses random effects we calculated a lot of interactions.
 

Dason

Ambassador to the humans
#27
Well it wasn't necessarily that I ran anything too complex. But looking at output and seeing if it seems reasonable for what they say they did. For instance knowing what the data is we can check degrees of freedom for certain things and see if they match what we should get if the person described things adequately. A lot of times things ... didn't check out the way they thought they should so I would have to look at their code more in depth to see what they actually were running.
 
#29
Could you please rank the programming languages in order and give each a score on a scale of 100? Particularly I'm interested in your opinion regarding C/C++, R, JAVA.
 

hlsmith

Less is more. Stay pure. Stay poor.
#31
Does your family members know that you have posted over 7 thousand times on this blog? Just from my perusal, it seems you may post ~ 3 times more than the second highest person (not taking length of membership into consideration). However, I will note that I do appreciate many of your insightful comments.

If your wife ever needs to get ahold of you or get your attention, all she has to do is post here in code (e.g., Post: Are sandwich estimators good [code for dinner)], then the reply may say something about Tukey and t-tests and she will know you want turkey sandwiches with Guiness.
 

Dason

Ambassador to the humans
#33
What are your views on users creating sock-puppet accounts? Can it ever be justified?
What an interesting question from somebody I can only assume is handsome and charming! I would say that it depends on the situation. For instance if somebody were to create a sock-puppet account mainly to make a joke and everybody realized that it wasn't a serious account - then it's probably alright. But... that's probably never happened right?
 

Dason

Ambassador to the humans
#34
Could you please rank the programming languages in order and give each a score on a scale of 100? Particularly I'm interested in your opinion regarding C/C++, R, JAVA.
I'm going to split up C and C++ (since they really are different languages) and remove C++ from consideration since although I know enough about it theoretically - I haven't done much coding in C++. To replace it I guess I'll toss in Python.

R > Python > C > Java (But really it depends on what you're doing)

R gets a score of Super Awesome / 100. It's very useful for statistics and takes care of a lot of stuff for you. Working with data is a breeze in R compared to other languages and there is package to do pretty much any statistical procedure you can think of. It's perfect for the type of work I do.

Python gets a score of Awesome / 100. There are very useful packages for working with data and integrating with R so doing real stats in Python is pretty easy. It's also much more of a general purpose language when compared to R so it's useful for a heck of a lot of stuff. Its syntax is clean and it provides some real nice syntactical sugar that makes doing certain things a breeze. It's definitely my scripting language of choice.

C gets a score of Useful / 100. C can be a real headache when debugging. Managing your own memory is useful for squeezing the most out of your code but it's really easy to mess up which can result in code that won't compile or something that ends up leaking memory like crazy. The actual language itself is incredibly small but you also don't get a lot of niceties either. If you're looking for speed it's hard to beat but... I try to avoid dropping down to C whenever possible.

Java gets a score of Meh / 100. It was my first real language so it has a special place in my heart. But. you. have. to. be. so. ****. verbose. about. everything. Have you ever seen a Hello World program in Java? Compare that to a Hello World program in R or Python and you'll see what I mean. Java does have nice things though (of course it does! otherwise it wouldn't be as popular as it is). The fact that you don't actually run the code on your machine is nice - you run everything in the java virtual machine which is what makes it possible to write cross platform code in Java so easy. There's also a library for EVERYTHING in Java. The problem is that this causes some people to not think too hard about their code so they end up importing a HUGE library just to do a simple simple task. This is partially what gives people the impression that Java is slow.
 

Dason

Ambassador to the humans
#35
What will you do when the next EMP burst gets rid of all the computers and calculators? Ever used a slide rule?
Well that would really screw up pretty much all the work I do. I actually have used a slide rule before though and I taught myself to use an abacus (just for the hell of it) a while ago.

But yeah... pretty much everything I do these days for work requires a computer. I guess I could always attempt to convert to more of a theory person... but I guess I'm just hoping an EMP doesn't go off anywhere in my vicinity.
 

Dason

Ambassador to the humans
#36
Does your family members know that you have posted over 7 thousand times on this blog?
Well this isn't a blog - it's a forum. My wife does know I probably spend too much time here. She doesn't know exactly how much more I've posted compared to others here but she does know that I have the highest post count.
 

bryangoodrich

Probably A Mammal
#38
Java gets a score of Meh / 100.
I agree about Java's verbosity. It does depend on what you want out of it, though. If you're doing OOP, it's amazing. Sure, not the greatest in some areas (and has actual structural flaws even the developer recognizes he would have done differently), but if I'm doing OOP, I'm probably looking at Java. Partly because I don't know C++ lol
 

vinux

Dark Knight
#39
What is your favourite movie? (non bot)
Who is your favourite Talk Stats member?
Why do you hate raptors?
Do you believe in god? (part of hypothesis testing)
what is your favourite food?

Now couple of stat/tech questions
who has most influenced you in your life? do you have any heroes?.
Do you have any memorable moments to share us? (exclude 100th level here)
I hope your research also align to gene expression data. What are the other area you would like to research on? ( this may after your Phd program).
 

Dason

Ambassador to the humans
#40
What is your favourite movie? (non bot)
Life is Beautiful I've only seen once but it was amazing. The Red Violin is another contender. I've seen that one quite a few times though.

Who is your favourite Talk Stats member?
Hamid. Definitely.

Why do you hate raptors?
They're evil vicious killing machines. Why would anybody NOT hate them?

Do you believe in god? (part of hypothesis testing)
Yes.

what is your favourite food?
Pasta. If I had to choose a specific kind it would probably be spaghetti but that's mainly because I made it a lot when I lived on my own. But my wife has a lot of Italian in her so her pasta is amazing.

who has most influenced you in your life? do you have any heroes?.
Statistically? Probably Dr. Hooks, Dr. Malone, and Dr. Deppa from Winona. They're the ones that initially got me interested in stats and kept me interested throughout undergrad. Since coming to Iowa State I would say that Dr. Hofmann and Dr. Nettleton have been my biggest influences (for different reasons).

Statistical heroes? He wasn't necessarily a statisticians but Richard Feynman is sort of a personal hero.

Do you have any memorable moments to share us? (exclude 100th level here)
Finding out I passed the PhD prelim! I saw on facebook that a few of my friends in the program found out their results but I hadn't heard anything yet so I emailed my advisor (this was on a Friday night). The next day I had to head to central Wisconsin to be a counselor at a camp. I hadn't heard anything back yet before leaving but once I got there I checked my email and my advisor had emailed back saying that I had passed. I was jumping up and down I was so excited! The only person around at the time was the janitor so I gave him a high five and he jumped up and down with me for a while. There is pretty much no reception there but I called Kelley right away and apparently the only thing she heard was "I passed" so she was really excited too.

I hope your research also align to gene expression data. What are the other area you would like to research on? ( this may after your Phd program).
My focus is gene expression data but a lot of that involves figuring out good ways to borrow information across genes. This could be used in different areas where a lot of data is collected. Finding ways to use the existing gene expression methods for other types of data would be interesting to me. I also really like teaching and I have a special place for stats education research. I've been a fan of Bayesian methods for quite some time so doing more with that kind of stuff would definitely be of interest to me.