FW: [R-sig-finance] import.data.rte in R?

[Dirk Eddelbuettel]

Jordi,

On Mon, Jul 05, 2004 at 03:44:48PM +0200, Molins, Jordi wrote:
> 
> Dirk,
> 
> when you say: "At this point I cannot release the code for it.", is it for
> copyright issues or for some other reasons? even though currently you cannot
> do it, do you think that you will be able to do it in the foreseeable
> future?

Yes, sorry, I wasn't very clear.  Both of those statements are true.  

As I did that development at work, it "belongs" to work and it is not
entirely my call if, and when, it gets released. "Intellectual property" is
increasingly seen as a competetive asset, and we are frequently reminded of
that, and even encouraged to think about patenting our work if 'suitable'.

I will need to lobby for my personal view that a connector package is
'merely' infrastructure, and that we'd be fine open source'ing it. But I
have have no idea how long that may take.  So 'future', yes; 'foreseeable
future' maybe not. I will certainly provide updates on that here.

> I am not a professional programmer (I am a trader), and if somebody that is
> proficient with the Bloomberg C-API could give me hints, I would be
> extremely grateful ... of course, I understand the problems of sharing
> propietary algorithms.

Your colleague Bernhard was also interested, and I have pointed him to the
(actually pretty decent) Bloomberg documentation and C API kit. Couple that
with some understanding one needs from hooking C code into R, and it is no
longer unsurmountable.  Maybe Bernhard and you need to find a sponsor to pay
for it inside DRKW, and then let a developer code it up for you. And maybe
you even get the bosses to release it ?

> I am already using the (D)COM port. It is pretty useful in the corporate
> environment. How would you use it for retrieving data into a database? you
> would call the import.data.rte, store the data in some array, and then use
> VBA code to send it to Access?

R can talk directly to many databases, RODBC helps a lot. The key would be
to get it into R first.

> About LIM: what is the advantage of LIM above Bloomberg or Reuters?

It's mostly complementary. LIM is for historical data, as well as the
ability to create / run scenarios in an almost normal language as in (NB:
untested, typing this from home)

	show TY: three day percent change in TY 
	when fedfunds is larger than previous value of fedfunds plus 0.49

See www.lim.com for more.

Dirk

-- 
White House officials praised the performance of the controversial 
new Diebold electronic voting machines, which successfully tabulated 
final results from Florida before a single vote was cast.
          -- Andy Borowitz, http://borowitzreport.com, 28 June 2004