MathML book

[Johannes Huesing <hannes@ruhrau.de>]

Dear ConTeXtuals,

Hans distributed some printed (double-sided) versions of the MathML
manuals on the Dante meeting so I took one and looked into it. I think 
I sent this message before but didn't see it on the list, so here the
gist:

The statistics examples are not quite correct in my opinion. They confuse 
between empirical values and their theoretical counterpart. The first 
example is correct. The second example is not incorrect but could be
formulated a bit more strongly by taking the expectation of the right 
and replacing the \approx by an equal sign. The third example is wrong
as the variance \sigma^2 is a theoretical value and the expression in
the middle its estimate. I would denote the estimate as s^2 or \hat\sigma^2. 
Moreover, I would place the exponent before the parenthesised argument
and not behind it, as \sigma^2 is a more basic value to us statisticians
than \sigma. 

For statistical expressions, I would like to see tags for expectation,
probability, estimate and conditioning (as in f(x|\mu, \sigma^2)). 
Symbolic algebraic programs could only use the expectation, the rest
would rather be syntactic sugar.

Thanks for the lift to Duisburg, Hans, hope you had a safe ride home!

Groet

Johannes
-- 
"Human Genome" may be a binary file. See it anyway?