Hi Folks,
I've been chipping away at two new libraries that I wanted to let people know about: A CFFI wrapper around the SciPy version of
Cephes and a
SQL-DF library modeled after the R package of the same name. I've also generated new docs (PDF, info and HTML format) for each lisp-stat library and checked them into their respective repos and updated the
lisp-stat website with new getting started guides and examples.
With these two additions I think lisp-stat is feature-complete enough (for descriptive statistics) to call it 'version 1.0'. Cephes gives us at least the same accuracy for statistical distribution functions as SciPy has, and with SQL-DF, SELECT and the built-in subsetting functions, data-frames are far enough along for useful work. So far, the
Vega-Lite plotting has worked quite well (see the
plotting examples), though a bug in declt prevents generating documentation.
A quick example of sql-df:
(ql:quickload :sqldf)
(use-package 'sqldf)
(pprint
(sqldf "select species, avg(sepal_length) from iris group by species"))
;; SPECIES AVG(SEPAL-LENGTH)
;; 0 setosa 5.0060
;; 1 versicolor 5.9360
;; 2 virginica 6.5880
I haven't requested these be added to quicklisp yet, as I wanted to get some feedback and use them a bit in my own projects to 'bake' a bit more. The repos are at https://github.com/Lisp-Stat.
Any comments or feedback are welcome.
Regards,
Steve