ICA update

Turns out that my implementation of step 5 from JMLR06 was wrong. The purpose of step 5 is to better visualize the graph. So in particular, the swaps need to be simultaneous row- and column-swaps.

The output of step 4, B-hat, already has all the information we want. So far, B-hat had always been lower-triangular. But that's because I've been feeding the variables to ICA in the order in which they were generated. When I feed them in a different order, B-hat is no longer lower-triangular, and swaps are necessary to make it so.

Since the W matrix (i.e. A^-1) has all zeros in the upper-right triangle, there is a unique row-permutation such that all the diagonal entries are non-zero. (We don't need to care about column-permutations, because we would get an equivalent answer... this is afterall the marriage problem, which is why we use the Hungarian algorithm)

There is a unique W row-swap which are non-zero.

awkward Portuguese writing

I have on two previous occasions blogged about the awkwardness of some kinds of expression in Portuguese. I've thought of an objective way to measure that.

* take an English text
* get a professional translator to translate it into Portuguese, as faithfully as possible
* among educated readers, measure how well they understand the text. Measure how many words of the words in the translation are known to them.

You will find that the translation has a lot of long, difficult words (unlike the original English text).

My hypothesis is that if you do the symmetric thing (Portuguese -> English translation), the same effect is not observed.

Conclusion:
* does English have a better set of words for talking about these things? does English morphology allow one to use the existing words more flexibly? (and among the morphemes that are translatable, are PT morphemes longer and more awkward, e.g. how EN "-ly" becomes PT "-mente")
* are EN->PT translators pretentious?


Here's the translation that inspired this post. Looking at it again, it seems that part of the awkwardness is due to the distance between formal and informal language in Portuguese.

It also seems to me that counterfactual sentences are awkward to express in PT "pintassemos" ( we were (subjunctive) ) takes 4 unpleasant syllables to pronounce. Likewise, the informal auxiliary "tivessemos" ("we had" (subjunctive)). This is even more painful because PT is syllable-timed (where EN is stress-timed). (Some have said that PT in Portugal (EP) is more stress-timed than Brazilian Portuguese (BP))