An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-sig-finance/attachments/20140314/4cad244f/attachment.pl>
Quantstrat help
4 messages · Noah Silverman, Joshua Ulrich
On Fri, Mar 14, 2014 at 10:30 AM, Noah Silverman <noahsilverman at ucla.edu> wrote:
Hi, Working on developing a complex portfolio and backtesting strategy.
From what I've read, the quantstrat package looks ideal for working
through the ideas. Two questions: 1) I have an indicator that is pre-calculated. (From a bunch of complicated functions and outside data.) I'd like to simply pass the vector of values into quantstrat, but can't figure out how to do that. The function to add a signal appears to only accept functions, not data. Is there a way for me to do this?
If it needed a function and there was no way to pass data, you could wrap your data in a function. datafun <- function(...) mydata But nothing *requires* that you call/use add.indicator(). Simply merge your indicator with your market data.
2) Along the lines of #1 above: I have my financial data directly from a vendor. It already has OHLC defined. Can I simply pass this to quanstrat? If so, what is the proper order/formant/naming of columns to make quanstrat happy?
That depends on the specifics of the strategy you're trying to build, which you haven't provided.
Thanks! -- *Noah Silverman, PhD* | UCLA Department of Statistics 8117 Math Sciences Building, Los Angeles, CA 90095
-- Joshua Ulrich | about.me/joshuaulrich FOSS Trading | www.fosstrading.com
An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-sig-finance/attachments/20140314/72b36a7f/attachment.pl>
On Fri, Mar 14, 2014 at 3:26 PM, NOAH SILVERMAN <noahsilverman at ucla.edu> wrote:
Josh, Your comment on the indicator makes sense. I already have it as a column in my dataFrame, so I guess that I can just pass that column name when creating a signal. The strategy is a *single asset* with OHLC each minute. I have a few custom indicators that are pre-calculated. Just need to test a bunch of signals and rules around those indicators and close price. Do I need to do anything special to have the dataFrame setup correctly for quantstrat?
Yes, you need to convert it to xts. ?applyStrategy says:
mktdata: an xts object containing market data. depending on
indicators, may need to be in OHLCV or BBO formats, default
NULL
The package has several demos, take a look at them to get a feel for
what you need to do.
Thanks! On Fri, Mar 14, 2014 at 8:52 AM, Joshua Ulrich <josh.m.ulrich at gmail.com> wrote:
On Fri, Mar 14, 2014 at 10:30 AM, Noah Silverman <noahsilverman at ucla.edu> wrote:
Hi, Working on developing a complex portfolio and backtesting strategy.
From what I've read, the quantstrat package looks ideal for working
through the ideas. Two questions: 1) I have an indicator that is pre-calculated. (From a bunch of complicated functions and outside data.) I'd like to simply pass the vector of values into quantstrat, but can't figure out how to do that. The function to add a signal appears to only accept functions, not data. Is there a way for me to do this?
If it needed a function and there was no way to pass data, you could wrap your data in a function. datafun <- function(...) mydata But nothing *requires* that you call/use add.indicator(). Simply merge your indicator with your market data.
2) Along the lines of #1 above: I have my financial data directly from a vendor. It already has OHLC defined. Can I simply pass this to quanstrat? If so, what is the proper order/formant/naming of columns to make quanstrat happy?
That depends on the specifics of the strategy you're trying to build, which you haven't provided.
Thanks! -- *Noah Silverman, PhD* | UCLA Department of Statistics 8117 Math Sciences Building, Los Angeles, CA 90095
-- Joshua Ulrich | about.me/joshuaulrich FOSS Trading | www.fosstrading.com