transformation MCMC

For reasons too long to describe here, I recently came across a 2013 paper by Dutta and Bhattacharya (from ISI Kolkata) entitled MCMC based on deterministic transforms, which sounded a bit dubious until I realised the deterministic label apply to the choice of the transformation and not to the Metropolis-Hastings proposal… The core of the proposed method is to make a proposal that simultaneously considers a move and its inverse, namely from x to either x’=T(x,ε) or x”=T⁻¹(x,ε) , where ε is an independent random noise, possibly degenerated to a manifold of lesser dimension. Due to the symmetry the acceptance probability is then a ratio of the target, multiplied by the x-Jacobian of T (as in reversible jump). I tried the method on a mixture of Gamma distributions target (in red) with an Exponential scale change and the resulting sample indeed fitted said target.

The authors even make an argument in favour of a unidimensional noise, although this amounts to running an implicit Gibbs sampler. Argument based on a reduced simulation cost for ε, albeit the full dimensional transform x’=T(x,ε) still requires to be computed. And as noted in the paper this also requires checking for irreducibility. The claim for higher efficiency found therein is thus mostly unsubstantiated…

“The detailed balance requirement also demands that, given x, the regions covered by the forward and the backward transformations are disjoint.”

The above statement is also surprising in that the generic detailed balance condition does not impose such a restriction.

 

another book on J.B.S. Haldane [review of a book review]

As I noticed a NYT book review of a most recent book on J.B.S. Haldane, I realised several other books had already been written about him. From an early 1985 biography, “Haldane: the life and work of J.B.S. Haldane with special references to India” followed by a “2016 biography “Popularizing Science” along an  2009 edited book on some Haldane’s essays, “What I require from life“, all by Krishna R. Dronamraju to a 1969 biography with the cryptic title “J.B.S.“, by Richard Clarke, along with a sensational 2018 “Comrade Haldane Is Too Busy to Go on Holiday: The Genius Who Spied for Stalin” by Gavan Tredoux, depicting him as a spy for the Soviet Union during WW II. (The last author is working on a biography of Francis Galton, hopefully exonerating him of spying for the French! But a short text of him comparing Haldane and Darlington appears to support the later’s belief in racial differences in intelligence…) I also discovered that J.B.S. had written a children book, “Mr Friend Mr. Leaky“, illustrated by Quentin Blake, Roald Dahl’s illustrator. (Charlotte Franken Haldane, J.B.S.’s first wife, also wrote a considerable number of books.)

The NYT review is more a summary of Haldane’s life than an analysis of the book itself, hard as it is not to get mesmerised by the larger-than-life stature of J.B.S. It does not dwell very long on the time it took Haldane to break from the Communist Party for its adherence to the pseudo-science Lysenko (while his wife Charlotte had realised the repressive nature of the Soviet regime much earlier, which may have led to their divorce). While the review makes no mention at all of Haldane’s ideological move to the ISI in Kolkata, it concludes with “for all his failings, he was “deeply attractive during a time of shifting, murky moralities.”” [The double quotes being the review quoting the book!]

Darjeeling shortage

When looking for Darjeeling tea in the recent days, I found out that the summer (or second flush) harvest has been very limited due to a 104 day strike in the region linked with the call for the creation of a Gorkhaland state and the separation from West Bengal. I remember discussing the issue last year in Darjeeling, with guides and taxi drivers, who were calling for the recognition of their cultural specificities, including the use of Nepali in local schools rather than Bengali. What amazes me a lot in this strike is the engagement of the tea garden workers, who have certainly few alternatives if any in terms of income sources. Unless I misread the situation and they were barred from attending the gardens… (Of lesser concern is the rise of the tea prices by a factor of ten for the highest quality leaves.) Hopefully, the tea gardens will recover from being left to grow wild for such a long while. And from workers leaving the region to find work elsewhere.