Vertigo

male masters

Yesterday, I received this wood plaque from the organisers of the San Francisco Marathon! Nice, as I had not expected them to deliver abroad. And with this surprising title of second place in the male masters. After checking on Google (and avoiding a long list of SM sites), it appears that this means the above 40 category, which I had never heard of before. (I would have preferred an above 50 category, obviously! Which appears to exist in some races as the Grand Master category, even better!) Looking further at some of the (running) links, it also appears that categories by weight have been introduced in some races… Not that I can expect to benefit from that further division!

MCqMC 2016 [#2]

In her plenary talk this morning, Christine Lemieux discussed connections between quasi-Monte Carlo and copulas, covering a question I have been considering for a while. Namely, when provided with a (multivariate) joint cdf F, is there a generic way to invert a vector of uniforms [or quasi-uniforms] into a simulation from F? For Archimedian copulas (as we always can get back to copulas), there is a resolution by the Marshall-Olkin representation,  but this puts a restriction on the distributions F that can be considered. The session on synthetic likelihoods [as introduced by Simon Wood in 2010] put together by Scott Sisson was completely focussed on using normal approximations for the distribution of the vector of summary statistics, rather than the standard ABC non-parametric approximation. While there is a clear (?) advantage in using a normal pseudo-likelihood, since it stabilises with much less simulations than a non-parametric version, I find it difficult to compare both approaches, as they lead to different posterior distributions. In particular, I wonder at the impact of the dimension of the summary statistics on the approximation, in the sense that it is less and less likely that the joint is normal as this dimension increases. Whether this is damaging for the resulting inference is another issue, possibly handled by a supplementary ABC step that would take the first-step estimate as summary statistic. (As a side remark, I am intrigued at everyone being so concerned with unbiasedness of methods that are approximations with no assessment of the amount of approximation!) The last session of the day was about multimodality and MCMC solutions, with talks by Hyungsuk Tak, Pierre Jacob and Babak Shababa, plus mine. Hunsuk presented the RAM algorithm I discussed earlier under the title of “love-hate” algorithm, which was a kind reference to my post! (I remain puzzled by the ability of the algorithm to jump to another mode, given that the intermediary step aims at a low or even zero probability region with an infinite mass target.) And Pierre talked about using SMC for Wang-Landau algorithms, with a twist to the classical stochastic optimisation schedule that preserves convergence. And a terrific illustration on a distribution inspired from the Golden Gate Bridge that reminded me of my recent crossing! The discussion around my folded Markov chain talk focussed on the extension of the partition to more than two sets, the difficulty being in generating automated projections, with comments about connections with computer graphic tools. (Too bad that the parallel session saw talks by Mark Huber and Rémi Bardenet that I missed! Enjoying a terrific Burmese dinner with Rémi, Pierre and other friends also meant I could not post this entry on time for the customary 00:16. Not that it matters in the least…)

San Francisco ½ marathon [1:26:32 – 15/7533]

Just completed the 2016 San Francisco ½ marathon. And this despite Air France’s best efforts to keep me out of it (and thanks to my son!), a long flight and hardly any sleep before the 5:30am start. This is the most brutal, unforgiving, relentless ½ I ever ran, with a significant positive differential of more than 700 feet (215 meters) and according to other runners with tracking devices possibly 0.3 miles extra (close to 500 meters). And fierce winds on the Golden Gate Bridge, both ways! So I am utterly flabbergasted by the outcome, which sees me arriving 15th altogether and first in the over 50 group [and third French!]… Even though I presume the training for Monte Rosa, the (unsuccessful) attempt at Monte Rosa, and the additional bike rides the previous month helped, although I did very little speed sessions. The race got hard the moment we started climbing, first a little hill near Fort Mason, then the slope to the bridge that truly slowed me down. Crossing the Golden Gate Bridge both ways was exhilarating, although I could not see one pile from the previous one, being in a cloud the whole time and Sun being not up yet. And although this section of the race was the most exposed, with no runner around to run in packs. Thankfully, there was one big downhill run around the 11th mile (that I passed in 1:11:11!) which helped me gaining back some time if no position, and facing the last climb, which seemed to last till the finish line in the Golden Gate Park… My overall time of 1:26:32 is surprising in itself if I account for the elevation: by Naismith’s rule, that would brings the time on a flat terrain under 1:20, a feat I only achieved once. And many other things are just weird in this race, from the 7500 runners I never saw, many of which finished over 4 hours, to the number of young winners (the third male runner is 17, the first female runner 20, a 14 year old came 22nd and even weirder the second female in the 5k is 8 year old!). And to the fact my ranking changed several times from 18th to 16th, 15th, 17th, and eventually 15th again.  [Congratulations to the organisers, by the way! The whole race was brilliantly organised with all kinds of amenities I had never seen before. And thanks to the supportive Erythrean taxi who took me from the airport and offered me a free ride back if I ended up in the top ten! It sounded like a joke at the time….]

off to Northern California

Before attending MCqMC in Stanford in two weeks, I will take some vacations in Northern California [really North!] with my family. Starting with the San Francisco ½ marathon tomorrow. So expect delays [as we already got stuck six twenty-seven thirty hours in De Gaulle airport thanks to a strike!] and mostly non-statistical entries on the ‘Og. And pictures.