Michael Pound

Sorry for only just getting around to this! This looks good, could you alter the commit to re-include the print("Loading Image") line before I merge?

Good spot! Feel free to put in a pull request if you'd like, or I can fix it. Sorry for the slow response, teaching and stuff! :)

Hi. There may be some python 2 / 3 issues! I wrote this in py2, I'm also pretty much only using Python 3 now. I'm happy for suggested fixes, but...

Hi Roel, Sorry for the slow reply - much work to be getting on with! I don't see why you couldn't do this, essentially you'd have to have the solve...

Hi! The data I used for this was apparently a count of the bigrams, trigrams etc. from the google books archive. Now I look though, I can't remember where I...

This is a very interesting problem :) But it's also turning out to be quite a challenge! Not having the facility to do this myself was something that bugged me...

That looks much better, I agree that (assuming ripple continues to decrease) 9-10 components should see very negligible ringing. How long do these effects take to compute on the GPU?...

Yes a typo :) It's no problem - Actually the last 6 months have been so busy with academic stuff it's nice to do something a little different! Optimising the...

Optimised 9 kernel: [7.393797857697906, 2.4737002456790207, -1796.6881230069646, 631.9043430000561] [13.246479495224113, 6.216076882495199, 3005.0995149934884, 169.0878309991149] [7.303628653874887, 7.783952969919921, -1058.5279460078423, 459.6898389991668] [8.154742557454425, 13.430399706116823, -1720.108330007715, 810.6026949975844] [8.381657431347698, 14.90360902110027, 1568.5705749924186, 285.01830799719926] [6.866935986644192, 20.281841043506173, 90.55436499314388, -59.610040004419275] [9.585395987559902, 21.80265398520623, -93.26089100639886,...

This one wasn't easy :) I ran the optimisation about 50 times for 10 components. 49 of them failed to find a decent optima, but one made it! Ripple is...