On reuters data in R.
I also had problems with those libraries, so hacked up something using three
tools:
I installed RExcel which allows you to pass data quickly between R and
Excel.
I installed RMySQL too, again allowing easy data passing between the R and
Mysql.
I used PowerPlusPro, the excel add in for 3000 Xtra, to set up a download of
reuters TS1 data to a spreadsheet just using the wizards, nothing fancy.
Using these tools, I set up a refreshing dataframe in R of the reuters data.
The basic flow was something like this:
1) I used PowerPlusPro (aka PowerLink) to fetch Reuters data to excel into
an excel named range.
2) an RExcel "rexec" call was made to run R code held in cells in the
spreadsheet.
2.a) remeber the rexcel function was run on each data refresh event
automatically, so powerplus pro was effectively synchronising the runs to
happen on update.
2.b) that R code put the named range of data into a tmp data frame in R, and
ran more R code to open a db connection, and move the block of new data to a
temp table in mysql using overwrite=T
2.c) it also executed a second RMySql call to start a MySQL stored procedure
that derived a delta from the tmp table, format it properly, and to insert
new rows data into my main instrument history table. That table had cols
named for working with R, things like row_names holding dates etc, turning
reuters "Last" into "Close" etc.
3) it lastly called R code to repopulate my main moving window of dataframe
data in R from MySql, which it then replotted before closing my connection
to the db
All this ran fast enough to work on min bars as they updated and was fun to
play with.
But to be clear - it was a terrible hack, and not a great way to work - if
there's a better way, I also need it.
Andrew
[[alternative HTML version deleted]]