## false value

Posted in Statistics with tags , , , , , , , , , , , on November 23, 2020 by xi'an

A very pleasant eighth volume in the Rivers of London series after a few so-so episodes! The relentless deadpan of Peter Grant is back full shape, the plot is substantial and gripping, new and well-drawn characters abound, and the story offers an original retelling of the Difference Engine. (Not that I have reservations about Gibbson’s plus Sterling’s 1990 version!) Including mentions of Jacquard’s loom, card fed organ automates, Ada Lovelace and Mary Somerville. Plus providing great satire on Ai companies with a hardly modified “Deep Thought” pastiche. Enjoyable all along and definitely a page turner that I read within three days..! And being strongly immersed in the current era, from the passing away of David Bowie to the dearful impact of Theresa May as home secretary. Presumably missing a heap of references to geek culture and subcultures, apart from Hitchhiker Guide to the Galaxy. And too many quotes to report, but some mentions of stats (“the Red Army had done a statistical analysis with demon traps just as they had with conventional minefields. The conclusions had been the same in both cases.” (p.50) and “Beverley climbed into the bath with a second-hand copy of Statistics for Environmental Science and Management” (p.69), which is a genuine book.) As often the end is a bit murky and a bit precipitated, but not enough to whine about. Recommended (conditional on having read the earliest ones in the series)!

## a journal of the plague year [grey November reviews]

Posted in Books, Kids, Mountains, pictures, Travel with tags , , , , , , , , , , , , , , , , , , , , , , on November 21, 2020 by xi'an

Read Evil for Evil, K.J. Parker’s second tome in the Engineer trilogy, published in 2009! Surprisingly, I remembered enough of the first volume for the story to make sense and I enjoyed it, for the same reason I liked Sixteen ways to defend &tc., namely for its attention to logistics and medieval industry taking over the muscle-display of standard equivalents, plus the self-demeaning attitude of most characters, again a welcome change from the standards! The pace of the story sometimes get bogged down, though.

Slowly cooked pulled pork with a hellish amount of red peppers, meaning I ended up eating most of it by myself over a few days. Tried cauliflower risotto, and liked it. Took my mom to a nice restaurant in Caen, À Contre Sens, after an oyster breakfast with her on the quays of a nearby Channel harbour, with a surprise lunch based on local (Norman) products. Finding hardly anyone in the restaurant due to COVID regulations made the experience even more enjoyable. And such a difference from the previous Michelin we sampled this summer!

Wasted hours watching the US presidential vote counting slowly unraveling, computing & recomputing from the remaining ballots the required percentage of Biden’s votes towards catching up, and refreshing my NYT & Fivethirtyeight webpages way too often. And remain fazed by an electoral system stuck in a past when less than 50,000 men elected George Washington.

Cleaned up our vegetable patch after collecting the last tomatoes, pumpkins, and peppers. And made a few jars of green tomato jam, albeit not too sweet to be used as chutney!

Watched the TV series The Boys, after reading super-positive reviews in Le Monde and other journals. Which is a welcome satire on the endless sequence of super-heroes movies and series, by simply pushing on the truism that with super-powers does not come super-responsibility. Or even the merest hint of ethics. Plus some embarrassing closeness with the deeds and sayings of the real Agent Orange. Among the weaknesses, a definitive excess of blood and gore, ambiguous moral stands of the [far from] “good” guys who do not mind shooting sprees in the least, and some very slow episodes. Among the top items, the boat-meet-whale incident, “Frenchie” from Marseille almost managing a French accent when speaking some semblance of French, and Karl Urban’s maddening accent that’s a pleasure to listen even when I understand a sentence out of two, at best.

## Les Creisses

Posted in pictures, Wines with tags , , , , , , , , , on November 20, 2020 by xi'an

## Wagram, morne plaine!

Posted in Books, Kids, pictures, Running with tags , , , , , , , , on November 20, 2020 by xi'an

Avenue de Wagram is one of the avenues leaving from Arc de Triomphe in Paris, named after a (bloody) Napoléonic battle (1809). This is also where I locked my bike today before joining my son for a quick lunch and where I found my back wheel completely dismantled when I came back!  Not only the wheel had been removed from the frame, but the axle had been taken away, damaging the ball bearing… After much cursing, I looked around for the different pieces and remounted the wheel on the bike. The return home to the local repair shop was slower than usual as the wheel was acting as a constant brake. I am somewhat bemused at this happening in the middle of the day, on a rather busy street and at the motivation for it. Disgruntled third year student furious with the mid-term exam? Unhappy author after a Biometrika rejection?

Not a great week for biking since I also crashed last weekend on my way back from the farmers’ market when my pannier full of vegetables got caught in between the spokes. Nothing broken, apart from a few scratches and my cell phone screen… [Note: the title is stolen from Hugo’s Waterloo! Morne plaine!, a terrible and endless poem about the ultimate battle of Napoléon in 1815. With a tenth of the deaths at Wagram… Unsurprisingly, no Avenue de Waterloo leaves from Arc de Triomphe! ]

## on completeness

Posted in Books, Kids, Statistics with tags , , , , , , on November 19, 2020 by xi'an

Another X validated question that proved a bit of a challenge, enough for my returning to its resolution on consecutive days. The question was about the completeness of the natural sufficient statistic associated with a sample from the shifted exponential distribution

$f(x;\theta) = \frac{1}{\theta^2}\exp\{-\theta^{-2}(x-\theta)\}\mathbb{I}_{x>\theta}$

[weirdly called negative exponential in the question] meaning the (minimal) sufficient statistic is made of the first order statistic and of the sample sum (or average), or equivalently

$T=(X_{(1)},\sum_{i=2}^n \{X_{(i)}-X_{(1)}\})$

Finding the joint distribution of T is rather straightforward as the first component is a drifted Exponential again and the second a Gamma variate with n-2 degrees of freedom and the scale θ². (Devroye’s Bible can be invoked since the Gamma distribution follows from his section on Exponential spacings, p.211.) While the derivation of a function with constant expectation is straightforward for the alternate exponential distribution

$f(x;\theta) = \frac{1}{\theta}\exp\{-\theta^{-1}(x-\theta)\}\mathbb{I}_{x>\theta}$

since the ratio of the components of T has a fixed distribution, it proved harder for the current case as I was seeking a parameter free transform. When attempting to explain the difficulty on my office board, I realised I was seeking the wrong property since an expectation was enough. Removing the dependence on θ was simpler and led to

$\mathbb E_\theta\left[\frac{X_{(1)}}{Y}-\frac{\Gamma(n-2)}{\Gamma(n-3/2)}Y^\frac{-1}{2}\right]=\frac{\Gamma(n-2)}{n\Gamma(n-1)}$

but one version of a transform with fixed expectation. This also led me to wonder at the range of possible functions of θ one could use as scale and still retrieve incompleteness of T. Any power of θ should work but what about exp(θ²) or sin²(θ³), i.e. functions for which there exists no unbiased estimator..?