My helicopter design class at the University of Maryland took a trip to the Frederick, MD airport last month for helicopter rides in a Robinson R-22. It was a great experience thanks to Advanced Helicopter Concepts. I’ve included a video here to share my experience. The video was recorded with a GoPro camera strapped to my knee. The high-pitched, low RPM warning occurs during the autorotation. Enjoy!
SUMMARY
Here’s a way to make sure you’re optimizing the writing of your code with an example in Fortran. It’s a neat, non-obvious trick to an engineer, but may be more obvious to a computer scientist. It involves the writing of a do loop where you are updating your value for an array, and instead of copying that array into a temporary variable, you simply use the programming logic to continue to use that information (location in memory) during every other time step. I was able to get 25% better performance for this simple change (and it scales). Not clear yet? Let me show you.
BACKGROUND
I’m a mediocre Fortran programmer but I’m learning new tricks with practice, challenges, and going to professor office hours. Taking Scientific Computing and High Performance Computing Systems right now is really increasing the strength of my programming skills. I hope to be an intermediate Fortran programmer by the end of the semester. Below is a sample from a code I wrote for a recent class project. The goal of the project was not to write a sequential version, so I feel comfortable posting a piece of the serial version.
CODE DESCRIPTION
The user is stepping in time from igen to gmax, calculating the neighbor value (‘naybs’) of a cell in the array ‘pop.’ Then, it updates the value of the cell in the array ‘pop’ for the next time step based on the neighbor values. But it does this update in ‘buffer,’ unnecessarily copying the data back to the array ‘pop.’ In this way, the loop proceeds without conflict. This works, but you may be sacrificing performance without even realizing it. I wrote EXAMPLE 1 below and then my instructor said “BUT WHY NOT DO IT BETTER?” and I went back to my computer and coded EXAMPLE 2.
EXAMPLE 1: Original Code
do igen = 1, gmax
naybs = 0
buffer = 0
! This loop finds neighbor values
do j = 2, y_limit + 1
do i = 2, x_limit +1
naybs(i,j) = pop(i-1,j-1) + pop(i,j-1) + pop(i+1,j-1)+ &
pop(i-1,j ) + pop(i+1,j )+ &
pop(i-1,j+1) + pop(i,j+1) + pop(i+1,j+1)
! Birth
if (pop(i,j)==0 .and. naybs(i,j)==3) then
buffer(i,j)=1
! Survival
else if (pop(i,j)==1 .and. &
(naybs(i,j)==2 .or. naybs(i,j)==3)) then
buffer(i,j)=1
! Death
else
buffer(i,j)=0
end if
pop(i,j) = buffer(i,j)
end do
end do
end do
EXAMPLE 2: Improved Code
do igen = 1, gmax
naybs = 0
if (mod(igen,2)==1) then
! This loop finds neighbor values
do j = 2, y_limit + 1
do i = 2, x_limit +1
naybs(i,j) = pop(i-1,j-1) + pop(i,j-1) + pop(i+1,j-1)+ &
pop(i-1,j ) + pop(i+1,j )+ &
pop(i-1,j+1) + pop(i,j+1) + pop(i+1,j+1)
! Birth
if (pop(i,j)==0 .and. naybs(i,j)==3) then
buffer(i,j)=1
! Survival
else if (pop(i,j)==1 .and. &
(naybs(i,j)==2 .or. naybs(i,j)==3)) then
buffer(i,j)=1
! Death
else
buffer(i,j)=0
end if
end do
end do
else if (mod(igen,2)==0) then
! This loop finds neighbor values
do j = 2, y_limit + 1
do i = 2, x_limit +1
naybs(i,j)=buffer(i-1,j-1)+buffer(i,j-1)+buffer(i+1,j-1)+ &
buffer(i-1,j ) +buffer(i+1,j )+ &
buffer(i-1,j+1)+buffer(i,j+1)+buffer(i+1,j+1)
! Birth
if (buffer(i,j)==0 .and. naybs(i,j)==3) then
pop(i,j)=1
! Survival
else if (buffer(i,j)==1 .and. &
(naybs(i,j)==2 .or. naybs(i,j)==3)) then
pop(i,j)=1
! Death
else
pop(i,j)=0
end if
end do
end do
end if
end do
I will say that you do have to write more code, but if that’s your hang-up, then you might not be very interested in performance in the first place.
More importantly, note the logic here. What I’m doing is identifying which time step is odd or even (by computing the mod of each time step), and then based on that result, I will update the next time step with a value from another grid. As far as I know, this can be done more elegantly in C by switching pointers. I don’t even know how to yet use pointers effectively in Fortran, so that might be possible here too. That would then give you the performance you desire, and the brevity that everyone likes. This result did increase my performance by 25%, which is dramatic if you are in the scientific computing world.
I note that this may be obvious to others, and they might even think that I made it harder on myself in the first place by doing something silly (which I did), but remember that I was taught this by multiple people, multiple times. Once I started to get an understanding of how data is held in memory, I started to make more advanced strides in programming. Hope this helps!
In this article I describe my favorite and must-have items for when I start my semesters in graduate school. A lot of what I list below can also be recommended for undergraduates.
What to expect:
Compiler
I am using the Intel Fortran Compiler, ifort, called Intel Fortran Composer XE 2011 for Linux. It’s version 11.1. You can find the compiler here, as part of Intel’s Non-Commercial Software Downloads. You can check the version of your ifort by supplying the command
$ ifort -v
The ifort compiler has OpenMP capability built in. OpenMP has a built-in ability to time the run that you are executing. One way that we can time the run natively in Fortran will also be shown.
At the beginning of my code, it looks like the following:
Example Code
program goodtimes
c$ use omp_lib
use your_modules
implicit none
double precision :: fstart, fend
[Declare other variables]
double precision :: ostart,oend
c Fortran timing
call cpu_time (fstart)
c OpenMP timing
c$ ostart = omp_get_wtime()
c Start of your meaningful code
c Middle of your meaningful code
c End of your meaningful code
c End Fortran timing
call cpu_time (fend)
c End OpenMP timing
c$ oend = omp_get_wtime()
write(*,*) 'Fortran CPU time elapsed', fend-fstart
c$ write(*,*) 'OpenMP Walltime elapsed', oend-ostart
end program
There are a few things to mention in the loosely-written code above. I wrote it a little Fortran 77-esque, where I started writing in the 7th column, and the OpenMP pragma is ‘c$’, ‘!$’, or ‘*$’. I used ‘c$’ above. In later versions of Fortran, use ‘!$’.
Notice that you must use the ‘omp_lib’ module in order to access the built-in ‘omp_get_wtime’. Otherwise, you will get an error. I strongly recommend making your start and end variables double precision. It doesn’t matter how you specify them as double precision, and I don’t necessarily recommend the way I did it above, but I just want to make it clear that you will have a better time with a double precision specification.
Note that cpu_time yields information about the CPU time (how long the CPU was working on your problem) and omp_get_wtime yields the wall clock time, such as the time that would have elapsed if you were timing the run from beginning to end with a very precise clock. I had a few runs for my application that showed the CPU operating at about 90% efficiency (where wall time is 100% of the total time). I recommend reading this post I did about profiling your code, so you can see which regions of your code are time consuming, and you can direct your OpenMP use in those regions.
Remember to include the ‘-openmp’ flag when compiling, and specifying the environment variable ‘OMP_NUM_THREADS’. I typically modify the ~/.bashrc file with a value and then source ~/.bashrc.
The rest of the code is self-explanatory. Read up on Fortran (or C) and OpenMP tutorials and other documentation for any additional information, or feel free to ask questions below. The C techniques are very similar and straightforward.
I had 3 goals at the outset:
But there are some advantages and disadvantages about this method in 1. because if you Print to PDF in Preview it will create a white border around all of your images in your new PDF. The alternative is to create a Service in Automator that will go through all your images and convert them to PDFs, then condense all PDFs to one PDF. I haven’t fleshed out this process yet, as I’m satisfied (i.e. unwilling to complain) with the white border.
I’ve been encumbered by the inability to create PDFs from multiple PDFs or multiple images for some time. Today I decided it would be an essential skill so I learned how to do it, and now you’re going to learn. It was another beautiful day where I felt justified in purchasing my Macbook Air. Note that the majority of this instruction was compiled from about 6-8 different websites (so you don’t have to go crazy over tiny misconceptions and errors), so I thought I would throw it all into this one!
Some helpful sites (but note that my write-up is a condensed version of all of these sites to save you some time):
Create PDFs from images
Multiple Page PDF from PDFs
Combine PDF files Service
Reduce file size through quality
Resolving Quartz Filter Issues
Multiple PDFs to PDF
After you’ve completed these steps, go to File and then Save! I saved mine as “Combine PDFs.”
These are the workflow steps within Automator to Combine PDFs.
Once you’ve closed Automator, test our your new Service. Highlight multiple PDF files and then Control + Click them to bring up several Finder options, and then scroll down to Services, where you should see Combine PDFs as one of the options.
Select your PDFs and ctrl+click, select Services, then select Combine PDFs and you’re done!
Let me know if it didn’t work for you in the comments and we’ll sort it out! Make sure you execute all the steps correctly before you accuse me of shenanigans!
Custom Reduction of Image Quality Using ColorSync Filters
I needed to reduce a PDF that I combined from something outrageous (like 200 MB to 600 MB) to something manageable (like 2-6 MB). How can we reduce the order of magnitude by 2 and still have recognizable quality? I had the same question. It turns out for practical purposes, if someone scanned a page of writing in very high TIFF quality (each image ~15 MB), they’ve overdone it. So let’s get it back to something more manageable.
I created a custom ColorSync filter, first. Later, this filter will be applied to my really long PDF as a postprocessing step.
Choose your type of Filter for the PDF.
Create your Filters by using ColorSync Utility.
Good luck!
Graduate school is not for everyone. This should be clear. It’s 1-3 years for a Master’s, and anywhere between 3-7 for a Ph.D. (based on your abilities and how bad you want it!) Many know immediately whether they do or do not want to attend. But for those who are on the fence for engineering, I offer guidance in the form of 10 reasons. My opinions may skew toward aerospace as that’s my background.
1. Grow-up
After four years (at least) of grinding away at some of the hardest problems you thought you could ever encounter, building a cascade of rocket science self-esteem, and emerging (by the skin of your teeth) with a diploma that will hold a special place…somewhere, eventually. Well, the four years spent between high school and college probably taught you many lessons, but don’t think for a minute you know it all. Graduate school raises the stakes. It will teach you valuable skills in time management, communication, engineering methods, etc., and you’ll be able to present these polished skills to your first serious employer.
2. Dictate your future
Having a master’s degree demonstrates to employers that you have added value because you took on additional responsibility and succeeded at that mission. Having a Ph.D. means you made a significant, unique contribution to human knowledge, and offers extremely rewarding career choices. Having only a bachelor’s degree isn’t that bad, but you’re part of a large group of similarly educated, bright, energetic, new-car-smell engineers who have to make a significant effort to stand out from one another. Unless you’re a networking phenom or you have your own blog…get comfortable dealing with resume stacks, phone interviews, and the brutal competition to earn your entry-level position. For those who already got a job with a bachelor’s, did you get your first choice out the gate?
3. Set an example
I have several younger siblings and a very large extended family. There are a lot of graduate students from all over the world, from smaller villages to huge cities, and we all remember where we came from. Friends and family back at home remember what we were like as children and they spread our inspirational stories. It is not only our duty as aerospace engineers to achieve the desired objective, performance, and safety of our operations, but to set an example for the next generation of engineers, so that they may be encouraged to follow us.
4. Broaden your perspective
Many universities encourage a global perspective. Your graduate work considers the work done by others around the world. The people you meet will come from all over, and many of them may work directly with you. A global (diverse) perspective is encouraged because the ability to see a problem from several different angles enhances your problem solving ability. Greater perspective can enhance your creative thinking and flexibility toward solving a problem. Aside from broadening your problem solving ability, graduate school will also give you additional time to meet people from other cultures and hopefully encourage your own personal development in understanding others.
5. Become a technical authority
Your value is not simply increased by the piece of paper representing your graduate degree, but by the technical knowledge and the advanced problem solving skills you now possess. When you have that knowledge, you are able to speak with authority on topics to guide decisions that will have significant effect on design and production (or at least, that’s how it’s supposed to work). If you obtain your Ph.D., your opinion will have much greater weight in your field, and you may be looked at when challenging questions arise within your area of study. Of course, you better know what you’re talking about!
6. Stand shoulder to shoulder with giants
Not only are the students and faculty in graduate school the smartest I’ve ever met, they are also making significant contributions to the subjects I care about most. It is a rare opportunity to interact with today’s most brilliant minds outside of the leading companies who can afford to hire them. You will not merely be observing your intelligent peers. You will be on the front lines, exchanging ideas and converging on problem solutions.
7. Improve your presentation ability
Without good presentation, the audience will not care about your work. During your master’s degree education you may be required to either defend your work as part of your master’s thesis. If you stick around for your Ph.D., you will be offered the opportunity at numerous conferences or poster presentations to explain what you’re doing. Even if you have confidence in your presenting ability, it will be a different challenge to meet the expectations of your audience in technical ability. You will also have great diversity in your audiences. You will be able to use the skills and knowledge gained in these presentations later in your professional or academic career. Here’s a helpful article on how to give a good talk, by Mike Dahlin.
8. Justifiable investment
This Georgetown study notes that 41% of aerospace engineers (surveyed among the labor force) have a graduate degree. Aside from this glaring statistic, you hardly need a napkin calculation of cost-benefit analysis to understand that graduate school is a justifiable investment. In many cases, the graduate school or a fellowship may pay for your education. You are additionally provided a very modest (but existent) stipend so that you can be a productive worker and not have to worry (much) about the bills. Often, you are getting paid to learn (which is also paid for). Take advantage of this very unique opportunity.
9. Money
We’re not looking near term, but long-term. When I started out as an engineering student in college, I had money as my lowest priority. Hell, I was proud of that, and couldn’t understand why people would get involved in engineering based on the $ sign alone. But I’ve come to understand that life isn’t cheap, and when it comes time to support a family, I definitely need to be a provider. Those with aerospace graduate degrees get an average earnings boost of 28%.
10. Freedom
You have the freedom to succeed and you have the freedom to fail. While this is always true, it is especially true in graduate school. There is so much flexibility in graduate school that you can choose to grind for days on multiple projects, churning out more results and ideas than you would have ever thought possible, or you can choose to let the hours go by without accomplishment. Depending on your specific field, you may have the ability to set your hours, work location, topic of interest, etc. The bottom line, however, is that you have the freedom to control the flow of your life, and you have the responsibility to ensure that the needs of your school, department, and advisor are met. Here are some productivity tips from Dr. Matt Might, a favorite blogger of mine.
Things you’ll need:
Basic steps:
error trying to exec `as': execvp: No such file or directory
Also note that I performed this installation on a Macbook Air.
gcc
Download and install Xcode by clicking this link, or by searching for it in the Apple App Store, where it can be downloaded for free (see image).
Xcode contains gcc
After you’ve downloaded Xcode, you’ll want to open it and agree to their terms of service. Then, you’ll want to navigate to the menu Xcode –> Preferences –> Downloads. Here you’ll see an option to download Command Line Tools (see image). Note that you’ll need a developer account at this stage, and I was redirected to their developer page where I had to fill out a form and create my account (using my existing Apple ID, where a lot of the form was already auto-filled).
Download Command Line Tools from Preferences --> Downloads
After you have successfully installed the command line tools, open your terminal and type something like:
$ which gcc
which should return the path of your gcc in /usr/local/bin. All of this should have been taken care of automatically.
gfortran
I mentioned at the beginning that I got an error when attempting to use gfortran on my machine before I’d even installed gcc. I found that gcc must be installed in order to use gfortran. But my gfortran installation went smooth because it’s very straightforward.
Download gfortran from this link.
After considering my hardware, I chose the option:
Mac OS Lion (10.7) on Intel 64-bit processors (gfortran 4.6.2): download (released on 2011-10-20)
The installation has a walkthrough that comes with the package, like many Mac installations. Straightforward and it should also work automatically. Then, open your terminal and type
$ which gfortran
and it should reveal that it was successfully installed in /usr/local/bin.
Happy programming!
