Physics update

We’re falling a bit behind on the physics course.  I was hoping we’d finish Chapter 3 this week, but one student has only done the problems and programs for Chapter 1 and read Chapter 2, and the other has done the problems and programs for Chapter 2 and only started the problems and programs for Chapter 3.  We will need to do at least 2 chapters every 5 weeks to finish on time, and I think that the later chapters will be a bit slower going than the early ones, so I really do want to try to do a chapter every 2 weeks for a while.

The lab write-ups have also not been happening.

Lab 3

The students have not yet analyzed the ball-drop clips we recorded 2 weeks ago. The assignment in Physics Lab 3 was

1. Analyze the ball-drop clips (in .mov format). For calibration, the distance we measured between the top and bottom stile of the file cabinet was 112cm. Since the ball was a few cm in front of the file cabinet, there may be some perspective error in using that measurement to calibrate the drop. Get Tracker to give you position, velocity, and acceleration plots. Use the fluctuation in the acceleration estimates to estimate the errors in the velocity and position measurements.
2. Write a Vpython program that simulates the motion of the falling ball including the initial pause before dropping, but not including the bounces.

I still want to see the results of that, but I suggest some modifications: using the higher-quality clip in 00179.MTS and using Tracker to do the modeling, rather than writing a separate Vpython program (though a Vpython program would still be ok, and would make it easier to later model the bounces as well).

Lab 4

I also want to see a write-up of the spring lab (Physics Lab 4), for which the students got the following measurements:

Spring Constant Data

Spring Letter
                         A       B       C       D       E       F       G       H       I       J       K

Dimensions
Relaxed Length (cm)    1.025   1.44    0.97    1.27    1.53    2.92    3.58    7.005   3.51    2.44    2.89
Coil Diameter (cm)     0.46    0.525   0.59    0.78    0.78    0.83    0.705   0.36    0.44    0.67    0.80
Wire Diameter (cm)     0.03    0.045   0.02    0.045   0.03    0.07    0.045   0.045   0.03    0.045   0.045

Forces
Stretch (cm)
0.5                    0.28    0.23    0.13    0.04    0.04    0.33    0.16    1.09    0.15    0.13    0.05
1.0                    0.47    0.34    0.23    0.06    0.06    0.48    0.26    1.57    0.24    0.18    0.10
1.5                    0.68    0.40    0.32    0.10    0.08    0.56    0.36    1.89    0.32    0.25    0.14
2.0                    0.87    0.57    0.39    0.12    0.11    0.68    0.47    2.37    0.38    0.31    0.19

Note 1: the forces from the force gauge are inconveniently reported in kg, not N, so need to be converted by multiplying by the scaling factor that the gauge presumably used (9.8 N/kg).  We could be more precise and use 9.80665 N/kg, but the measurements are not accurate or precise enough to need the extra precision on the strength of the gravitational field.

Note 2: the students did not record the number of turns on each spring, which would provide a good check on the wire diameter, which was measured with a micrometer calibrated in inches.  The reported number is probably the result of a conversion, but the raw reading from the micrometer should have been recorded.

Correction: I have been informed that the students did not use the micrometer to measure the wire gauge, but the plastic calipers, which measure in metric units, but may not be very accurate for such small measurements.

I counted the turns myself:

                         A       B       C       D       E       F       G      H      I       J       K
turns                   20      27      18.8    26      31.5    37.9      51     98.5   65      39.8    31.2

Further corrections

I also remeasured the lengths, using new stainless steel calipers that just arrived this week (so the students couldn’t have used them).  It looks like spring K had been mis-measured (or, more likely, a “1” -> “2” typo).

                         A       B       C       D       E       F       G      H      I       J       K
length(cm)              1.025    1.415   0.985  1.305   1.585    2.965  3.60    7.03  3.45    2.40    1.935

Estimating wire gauge from these measurements, I get 4 different wire sizes:

                         A       B       C       D       E       F       G      H      I       J       K
wire diam(cm)            0.05    0.05    0.05    0.05    0.05    0.08    0.07   0.07   0.05    0.06    0.06

End Corrections

wire diam (cm)

End Corrections

For each spring, determine the following:

• Is the force to extend the spring linear with the increased length for each spring?
• If so, what is the spring constant (in N/m)?

Assuming that all 11 springs are made from the same material (which seems to be the case), come up with a model (a formula) that estimates spring constant given just the dimensions of the spring.

After looking at the data, I’m not sure that these measurements are accurate enough to be able to derive a good model from them. We may be able to do something by applying physical reasoning, the way we did for determining the effect of putting 2 identical spring in parallel or in series, but I don’t think that just doing arbitrary fits to the data will help us much.

Priorities

1. The highest priority should be on finishing the exercises and programs for Chapter 2, so that we can catch up.
2. Second priority should be on analyzing the data for the springs for Lab 4, so that we can discuss the data next week.  If it does not match our expectations (that is, if some of the springs seem non-linear, or if a couple of springs seem to be very different stiffness from what we predict), then we may want to redo some of the measurements, either to confirm that we measured and recorded the data correctly or to fix errors in our record-keeping. The formal writeup can wait until we have discussed the data (and redone measurements if necessary), but the analysis of the initial data should be done by Nov 4.
3. Third priority is reading Chapter 3 and finishing the problems and programs for Chapter 3. We will discuss the chapter at our next meeting, after clearing up any problems with Chapter 2 and discussing the spring data.
4. Fourth priority is playing with Tracker to try to extract data and model the ball drop.  Note that Tracker’s velocity and acceleration computations assume smooth changes in velocity, not the acceleration spikes you get from bouncing a stiff ball like a ping-pong ball.  I now have a copy of the source code for Tracker, and I am looking into providing better velocity and acceleration computations for collisions whose duration is less than the time between samples, but for now we’ll have to use the data that Tracker currently provides.
   The data to play with is that in 00179.trk from the HD clip in 00179.MTS, which is of higher quality than the previous clips, though I’m sure we could do better with a brighter light and a dark ball in front of a light background.  The clip has at least 2 ball drops in it, but the trk file only tracks the first 7 bounces of the first drop.  Try creating a “dynamic particle model” that matches the fall until the first bounce.  (It is a bit messier to create a dynamic model in Tracker that matches multiple bounces, though a piecewise model could be built with the “if(cond, then-expr, else-expr)” construct.)
   It might be worthwhile to try to create a new “Point Mass” object tracking the bounces of the second drop, setting up the coordinate frame so that the floor is at 0 and the ball starts on the positive y axis (rather than the negative x axis, as I did in tracking the first drop). You need to use Video-Filters-Deinterlace to get clean enough images for the Autotracker to work.
5. I never got writeups for how the ultrasonic range-finders work.  I still think that this is a valuable writing exercise, needing some on-line research.
6. Lowest priority is writing up what was done  in the Tracker lab with the ball drop.  We may let this one slide, since the main goal was familiarization with Tracker for later labs, rather than modeling physical phenomena.

Coming up

It might be valuable for us to try analyzing the low-speed collisions in Newton’s 3rd Law (or How to Make Effective Use of Video for Instruction)  More specifically to compute 2D momentum for each cart before and after collisions in the video clips for each collision.  Since we are not given the masses of the carts, we may have to estimate them from what should be happening (unless the long video at the beginning of the post gives us the masses—I didn’t watch it).

Since we do not have carts with hoop-spring bumpers like the ones in the video, it would be difficult for us to do this lab ourselves.

DARPA Shredder Challenge

The Defense Advanced Research Projects Agency wants to improve spycraft by making it easier to reconstruct shredded documents automatically: DARPA Shredder Challenge. The puzzle itself is mildly interesting: reassembling scanned images of shredded documents is not a trivial problem, but for small enough problems some simple brute-force approaches look like they could be used.  Bootstrapping those approaches to larger problems is an interesting challenge that may best be solved by huge computing resources like Google has.  There are obvious applications in archæology and for banks and businesses that mistakenly shred important documents, as well as the spycraft that DARPA cares about.

But DARPA are not willing to spend any money to achieve this goal, so they are attempting a prize competition.  This has worked well for them in the past (with the autonomous vehicle challenge) and it worked for Netflix (improving their movie recommender), but I think that this time DARPA has gone too far in the cheapskate direction.  This time they are offering only a $50,000 prize. That will pay for one grad student or postdoc for a year—if you win.  It probably wouldn’t even pay for the computer time needed to solve the biggest problems. Even Netflix, with a simpler problem, knew enough to offer a million dollar prize.

They may get a few hobbyists or students interested in the problem, but is unlikely to interest many professionals—the low odds of winning combined with the tiny prize make the expected value of entering the contest too low to invest more than a few hours in.  Perhaps they think that there is a system already out there (perhaps hidden in one of the government’s many secret agencies), but not advertised, and they are just trying to get their hands on it. $50k may be just enough to bribe a low-level flunky to release something illicitly (I wouldn’t know, having never bribed anyone and never been offered a bribe), but it isn’t enough to fuel development.

 

News from Stanford CS

I got the annual alumni letter from Stanford Computer Science today, and I noted a few interesting factoids, along with the piles of rather uninteresting news about people I didn’t know:

• Stanford has seen an 83% increase in computer science undergrads over the past three years, which they attribute to their new undergrad curriculum, which started in 2008–09.
• The CS program is swallowing the Computer Systems Engineering program, making it a track in CS.
• The CS department estimates that 90% of all Stanford undergrads take at least one CS course.
• The admission rate to their PhD program is about 10.3% and to their MS program is about 18.4% (but their co-terminal masters has a 96.7% acceptance rate, which they attribute to clear guidelines about who will be accepted, so few futile applications). Both grad programs have higher acceptance rates than Stanford undergrad admissions (7%), but CS grad programs are hard to get into at Stanford.
• The PhD program has started doing 3 lab rotations in the first year, as is common in some other fields (like the UCSC bioinformatics program), but Stanford thinks that they are the first pure CS department to do rotations.
• Stanford has eliminated their comprehensive exams in CS but not put in place any course requirements.

I’m not hugely impressed with the growth in undergrad enrollment: UCSC computer science grew 140% in 3 years without swallowing computer engineering, which also grew 122% (yes, both more than doubled from 2007–08 to 2010–11). [planning.ucsc.edu/irps/majors/2010/Historical_3QtrAve_UndergraduateDeclaredandProposedMajors(HC).pdf]

Having 90% of undergrads take a CS course without it being a requirement is pretty impressive.  But Stanford does require a general education course in engineering or applied science, and CS may be providing the easiest of such courses, so their success here may be from offering some courses seen as super lightweight ways to meet a requirement, rather than from real interest in learning anything.  I wonder if they have ever bothered to check that, or if they even care.

I’m pleased to see that Stanford is finally doing lab rotations in CS for the PhD students.  I did that on my own (thanks to fellowship funds) when I was a grad student at Stanford, but many students were left to flounder or work for the first lab that would take them. I generally spent more than a quarter on each project I worked on, but a more disciplined quarter on each would probably have gotten me through the program faster.  Not that I was in any hurry—being a Stanford CS grad was a great experience, and I would have stayed longer if my funding hadn’t dried up.

Stanford’s elimination of the comprehensive exam, while continuing their no-courses approach to grad school tells me that they have gone into depth only at the expense of breadth.  What I valued about my time at Stanford was the enormous breadth of research to be involved in.  They seem intent now on producing very narrowly focused researchers, which will probably mean that future Stanford grads will be less adaptable and thus less valuable in startups or faculty positions.

 

 

Oregon’s National Career Readiness Certificate

I just found out about Oregon’s National Career Readiness Certificate, which appears to be a test of high-school level competency in reading and math.  It appears to be replacing the high-school diploma in Oregon as certification of minimal competence for jobs seekers.  The number of people with such certificates is still small (under 11,000), but that is about 22% of a high school cohort in Oregon (or a 1/3 of a graduating class, since only 2/3 of an Oregon high school cohort graduates in 4 years).  Note: Oregon is doing even worse than California, which has a 4-year graduation rate of 3/4.

Because high school diplomas have gotten almost meaningless in most states, with graduation standards so low that employers can’t count on much from students who have them, having an independent certificate of minimal competency could be valuable.  The certificates are probably even more valuable for people who dropped out of high school, since the lack of a high school diploma is normally seen as proof of incompetence, when that need not actually be true (students drop out for many reasons, not just incompetence).

I normally think and write more about the educational needs of students at the top of the ability spectrum, because I teach graduate students in a challenging field and my own child is highly skilled, and so I see the holes at the high end most clearly.  I do care, however, about students at all levels of ability, and I think that the career readiness certificate may offer some hope of decent employment for those who fail to finish high school—certainly it seems to be attracting more people than the GED certificate does. I wonder whether California will implement a similar system, or buy into Oregon’s claim of “national” career readiness and join them in a shared certificate.

40% increase in female CS majors—not as good as it sounds

According to a press release this week from UCSC (Baskin School of Engineering promotes increased participation of women in computing), UCSC has had 40% increase in the number of women majoring in computer science in the past 2 years.  Sounds great, doesn’t it?  But it is just spin.

The article does not say how big the increase in the number of men majoring in CS  was over the same period.  According to the 3 Quarter Average of Undergraduate Declared and Proposed Majors (historical, HC), the total CS majors went from 185.7 in 2008–09 to 294.9 in 2010–11 (fractions result from averaging over 3 quarters and from double majors).  That is a 59% increase, so it seems like the fraction of women in CS has been dropping.

Indeed, this quarter, Undergraduate Majors by Gender (HC) shows 30.0 women and 259.5 men, or about 10.4% of the majors being female. Even adding the proposed majors (many of whom will disappear to easier majors) only gets the numbers to 57.0 and 467.0 (10.9% female). This is not quite the lowest ratio of women of any of the engineering departments at UCSC (bioinformatics 29%, computer engineering 10.6%, electrical engineering 8.5%, bioengineering 26.5%, information systems management 16.9%), but is certainly not a number to brag about.  Actually, none of those numbers are anywhere near what they should be. (The bio- ones look relatively good, until you compare them with MCD bio, which is 59% women.)

Unfortunately, I can’t find on the planning website historical information about majors by gender, so I can’t do direct comparisons of the same measurements, but it looks to me like the engineering disciplines are getting worse gender imbalances, not better.

If the article had been about the growth in CS majors (thanks mostly to the game design major), it would have been a good, honest article. The accomplishment of growing CS enrollment at a time when many colleges are seeing shrinking CS enrollment is a story worth telling. But trumpeting the growth in female computer science majors, when their fraction of the CS majors is probably shrinking, is just sleazy advertising.  I’m once again ashamed to be associated with a university that will stoop so low.