I’m back at work on my SLAM based laser scanner. Failure is not an option. Yet it feels like there is a real risk of it.
One of the steps in the process is to displace all the inertial unit measurements by a small error term in order to minimize the error in the correspondences.
More simply, we have a matrix A of height m and width n (m>n), a column vector b of height n, and we want to fill in the column vector x of height m such that A x = b is almost true.
There is no exact solution, so we look for a least squares answer, where (A x – b)^2 is minimal.
Luckily, there is a function scipy.linalg.lstsq() which does the job.
Let’s consider a simple example of a 3×2 matrix:
I just got myself a new laptop and installed Ubuntu-Linux on it. Scares the hell out of me the speed with which I got it up and running. I am now lost in a sea of code. It’s like walking into a public library after you’d been out in the sticks for a month with only two dog-eared issues of the Reader’s Digest to keep you company. There’s almost too much here. I want to read all of it. And any book or manual you do pick up and spend an hour with means there’s another ten thousand you’ve not picked up that you should have been reading.
Anyways, while doing my apt-cache searching stuff for stuff, I noticed stimfit – Program for viewing and analyzing electophysiological data show up in the search for scipy.
It appears to take datasets of electo-potential readings from a single neuron at every tenth of a milisecond and then fit exponential decay curves [the thick grey line] to selected sections from the (negative) peak to the baseline.
A bit like a temperature sequence, eh?
Oh, and it has a funky Python shell built into it to help you automate the analysis functions. What’s not to like?
Wednesday, October 22nd, 2014 at 3:02 pm - Weekends
How does anyone hold down a proper job?
I just got a “Hello world” program working out of a pair of Jeenodes kicking around in the cardboard box left over from the Housahedron project before they migrated to Berlin. Of course, there was no documentation for how to plug in the interface into the Jeenode, and I had to get Adrian’s help.
To make up for my disorganization with the data collected at my house (in that it got lost, was not frequently sampled enough and didn’t happen over the winter), I got this lovely temperature sequence from megni to analyze for my exponential decay theory which took a reading inside their cottage every 60 seconds.
My theory is that by fitting exponential decay curves to the data I would get some invariant values relating to the fabric of the building that would change when you improved its insulation characteristics (eg draught-proofing a window).
The first step is to chop of this data into the sections where the temperature is dropping down. It took a while to get some working code, but it came like this:
gw = 30 # half an hour sampleseqs, sampleseq = [ ], None for i in range(gw, len(samples): vd = samples[i-gw] - samples[i] # positive if past temp higher if vd >= 0: if not sampleseq or vd >= mvd: # restart seq at bigger difference sampleseq = samples[i-gw:i] mvd = vd sampleseq.append(samples[i]) elif sampleseq: sampleseqs.append(sampleseq) sampleseq = None
Suppose we have a temperature sequence like this one gathered from last September in our kitchen when we started to put the fire on in the late evenings. The central heating wasn’t on yet, so there’s no second temperature “bump” in the mornings. (The three bumps in the first peak are the three logs we put on the fire.)
You get an appreciably square looking graph by plotting the data as units of an hour in X, and units of a degree centigrade in Y, so a 45degree slope would represent a 1 degree difference per hour, which is the right scale of change in our environment.
My immediate observation from the first moment I saw such a temperature sequence (roughly in the middle of last summer when we visited megni in North Wales) was that these are exponential decay curves.
The trick is to find them and fit them.
Tuesday, October 14th, 2014 at 10:42 am - Machining
Friday evening I made a visit to the Berlin Fab-Lab open day. They’ve got a heck of a lot of 3D printers in a small space. All kinds colours and materials, from brittle and hard to rubbery plastic. I think they also build their own kits. (I asked them if they’d heard of the Autodesk Spark, and they hadn’t. It’s great to be out in the big wide world!)
But just as they find in DoESLiverpool, the 2D laser cutter gets the most use, because we can design things in 2D for a fraction of the effort.
The weekend was blighted by a desperately bad headache which was entirely unlike a hang-over that confined me to a dark room. (Hang-overs tend to release at around 8pm the following day for me.)
On Sunday afternoon I started to do some work.
Firstly, I discovered that all my laser scanning data was lost on the Autodesk computer which I gave back last week for disposal, so I can’t work on that software till I get some more, probably by going to Bristol and making the device work again.
Then I discovered that almost all my temperature sequence data was also lost on that same computer — although I do have 6 days of records from some time last September before the cold weather properly set in and we got some actual useful data. Whether I can find it anywhere on an SD card, or I have to collect some data all over again with a new Arduino set-up of my own making will have to wait till I get home.
For convenience, I’m putting the maths of exponential decay curves into a separate blogpost.
Here’s a picture of some not very spooky pumpkins in the local supermarket.
Also, I spied a poster for the newly opened Happylab Salsburg. Might be one for a drop-in next expo.
Maybe these are like computer clubs were back in the 1970s. It’ll all make sense in hindsight one day.
Thursday, October 9th, 2014 at 3:43 pm - Machining
I’m casting around for some little long term geometric projects which I could be good at. I’m very bad at the sysops stuff and compilers (which seems to be a breeze for every other hacker in the world). Plotting the geometry which you have calculated is also a drag.
For me, twistcodewiki does it all.
I followed the instructions to compile OpenVoronoi on this very small under-powered linux netbook I have kicking around, and got twistcodewiki to work. Here is me entering a polygon and plotting a voronoi structure from it:
The code is as follows:
Doesn’t matter in the long term. I am still absolutely certain that this will be a consumer product, like a TV or a microwave oven. Every house will have one.
I think I am coming down with my winter cold. Lethargy and crawly skin. The train journey from London was a lot tougher than it should have been. I am now back on my old pre-Autodesk laptop. There was a bit of panick getting the Webgl drivers to work so I could run my essential twistcodewiki system. Then I chatted with Mr. Heeks of HeeksCNC and made an inspection of the state of the art in the Open Source CNC world. It’s waiting for the time when geeks own their own machine tools so they are not only in the hands of professional engineers who don’t like to program.
The whole HeeksCAD UI is in C++ wxWidgets and too complicated for me to compile (especially as I’ve uninstalled the compilers), but it works by writing out a short Python program that it runs in a separate executable which links against area_funcs, libarea and nc (for the post-processor) in ways that rely on the global namespace (ahem).
Anyways, I’ve just managed to get it to operate in my WebGL system, so the point is proven. Don’t know where it leads, but one of the intentions is to get a computer system that drives a machine tool directly in a tight feedback loop in a similar manner that it would drive a car — ie not by generating all the motions of the steering wheel off-line and simply replaying them from a file. That would be some innovation that I cannot ever see happening in the real world, because the idea is not an incremental change strictly limited to one of the hardened silos in which this field of engineering is strictly divided.
Other projects not touched on yet are my laser scanning software (I’ve lost my sample files), the undemocracy scraper, tunnelx cave surveyor, triangle machine tool kickstarter, and some android phone apps.
This CEO of Autodesk has taken everything I’ve made in the field of machine tool software in return for some stock options that I cannot cash and a crushing of my spirit. Nevermind the price; I need to get out before I lose my mind.
This is the time to create negative publicity for myself. The teflon was scraped off years ago. Burning your bridges is alive in the hackspace.
Is it inevitable that any large hierarchical organization will deteriorate into a pool of grease and bureaucracy around one exquisite bubble containing the CEO?
Since he possesses the sole discretionary power of patronage and the gift of money, the sapiens do gather round and stare as if at camp fire. No one is permitted to spit.
The first thing that happened when I became CEO was I got much funnier and smarter… I would say there’s a part of it that you can resist and resist and resist, but there is a part of it where there’s a whole organization or a world around you that looks out for you in a weird way.
Still, it must be tough at the top of a creaking ten billion dollar thirty year old software corporation, raking in the multi-millions of dollars that awarded during this post-financial crisis corporate CEO pay bubble era.
Meanwhile, paradigm-shifting software products are popping up left and right out of nowhere within miles of your downtown San Francisco offices, and nothing happens on your patch. The magic has gone. You can pay over the odds for any tech start-up you like, but you’re not fooling anyone. It’s simply retail therapy.
Friday, October 3rd, 2014 at 11:55 am - Kayak Dive
Aside from having to get up at five in the morning and Becka not skiving the day to come along on the boat, it was a perfect trip out.
Liverpool is starting to look a lot like the Esbjerg did the first time I caught the ferry over to Denmark in 2003 and saw spinning wind turbines everywhere.
What amazing kit — the original gopro camera still hasn’t broken. I record these videos for the same reason I write phone numbers down in a book or keep a trip log: I can’t remember things well enough.
The water was pretty warm at 16degrees. I wore all my layers under the drysuit anyway and barely felt a thing swimming around. It was like a dream. I didn’t take any lobsters, but the others did. The conger eels stayed in the cracks while the tompot blennies came out to play.
The second dive was on the Calcium where there was one large cod and more starfish than grains of sand.
There was a surprise party in the evening, which I thought was for my birthday, but wasn’t.