https://github.com/cran/quantmod
Tip revision: 65a1956a0c8ce499f7ce2197e1cdcee76bfdb1ed authored by Jeffrey A. Ryan on 09 June 2008, 00:00:00 UTC
version 0.3-6
version 0.3-6
Tip revision: 65a1956
WISHLIST
********************************************************************************
********************************************************************************
quantmod wishlist: as of Mar 21, 2008
********************************************************************************
********************************************************************************
- Documentation (web and vignettes) :
1. charting and indicator overlays!
2. building models
3. testing models
4. extending quantmod
5. package/inteface preferences
- Move to R-forge for devel tools [done]
- getSymbols.skeleton, buildModel.skeleton, addTA.skeleton
- convert to xts from zoo throughout. Simplify data methods [started]
- try to remove dependence on fSeries and its (etc). Through 'xts'???
FUNCTIONALITY:
- getFundamentals.google and getFundamentals to download data
from google. Possibly add arbitrary downloads from finance.yahoo.com
using http://www.gummy-stuff.org/Yahoo-data.htm formats
[ renamed getFinancials (google) done - jar ]
- tradeRule and tradeLog
- move all quantmod.OHLC to c('OHLC','xts','zoo') format - eventually move
to new package all together [ xts is ready for move - jar]
CHARTS:
- add /demo script
- add addHolidays/Weekends function to add in missing data with NA [see fillIndex beta in 'xts']
- liveChart function - used in conjunction with RT data feeds [SEE BELOW]
- add x to reChart to allow for updated/different data, which may
be used in conjuction with a RT feed callback such as in the IBrokers
package
- add 'periodicity' argument to chartSeries
- dropTA function to remove TA indicators applied [moveTA and swapTA added as well: 2/16/2008 - jar]
- add subset argument to chartSeries (et al) and buildModel function to pass
along to underlying [.xts method [done, along with zoom & zoomChart - jar ]
- mask lagged indicator days - use full dataset for calculation and a
seperate subset for actual plotting. (using above [.xts method) [ done see above - jar ]
- multiple series in one line chart, or in one window - possibly
mchartSeries, mbarChart, mlineChart, mcandleChart plus multiple
series normal command interpretation
** OR **
partition device in n-regions and have chartSeries.chob check
- HSV color schemes from vcd or colorbrewer
- finish TA integration with TTR
x additional charting tools, like seriesHi, seriesLo, lines, points,...
x allow for indicators to be drawn under series - BBands, Envelopes...
- performance charts:
- periodReturn - chartSeries method(no!) to handle
- modelResults
- integrate with PerformanceAnalytics
TESTING:
- regression testing suite, via RUnit.