indicator_attribute: Add attribute to indicator timeseries of stocks

In chriszheng2016/zstmodelr: Stock Factor Models for China Stock Market Investment

Description

Compute attribute values of stocks and add them to indicator timeseries.

Usage

attr_indicators(
  ts_indicators,
  new_attr_def,
  parallel = getOption("zstmodelr.common.parallel", TRUE)
)

attr_indicators_indcd(
  stock_db,
  ts_indicators,
  parallel = getOption("zstmodelr.common.parallel", TRUE)
)

attr_indicators_trdstat(
  stock_db,
  ts_indicators,
  parallel = getOption("zstmodelr.common.parallel", TRUE)
)

Arguments

ts_indicators	A dataframe of indicator timeseries.
new_attr_def	A function to compute attribute value for stock indicator.
parallel	A logic to determine whether to use parallel processing. Default TRUE means to use parallel processing.
stock_db	A stock database object to get data.

Details

attr_indicators is a working horse behind attr_indicators_indcd,attr_indicators_trdstat, which is mainly used to develop customized attribute function to add attribute to indicator timeseries.

attr_indicators_indcd, attr_indicators_trdstat, etc. are used to add predefined attribute to indicator timeseries by using data from stock_db.

There are two way to add attribute to timeseries of indicators of stocks:

• add attribute directly: use attr_indicators_indcd, attr_indicators_trdstat to add attribute directly. The pros is easy to use, the cons is slow since it have to load data from stock database each calling. The method is mainly used to add one attribute to one indicator timeseries.

• add attribute indirectly: use ind_attr_def_indcd, ind_attr_def_trdstat and modify_indicator. The pros is that it can be faster by dividing defining attribute(load data from database once) and modifying indicators (can modify repeatedly by using same attribute definition). The cons is that it is more complex than direct method. The method is mainly used to add same attribute to multi-indicators timeseries.

Value

A dataframe of indicator timeseries with new attribute if succeed, otherwise NULL.

Functions

• attr_indicators: add customized attribute to indicators timeseries of stocks.

• attr_indicators_indcd: add attribute of industry code to indicator timeseries of stocks.

• attr_indicators_trdstat: add attribute of trading status to indicator timeseries of stocks.

Examples

## Not run: 

# create a indicator from vars dataset.
ts_indicator <- create_indicator(ds_def_vars,
  ind_def_fun = ind_def_fun,
  debug = debug,
  date_index_field = "date",
  key_fields = key_fields,
  parallel = parallel
)


# --- add attribute directly ---

# add attribute of indcd
ts_indicators_with_indcd <- attr_indicators_indcd(stock_db,
  ts_indicators = ts_indicators
)

# add attribute of trdstats
ts_indicators_wide_with_attr <- attr_indicators_trdstat(stock_db,
  ts_indicators = ts_indicators
)

# --- add attribute indirectly ---

# create attr defs by using data from stock
new_attr_indcd <- ind_attr_def_indcd(stock_db)

# use attribute def to modify multi-indicators
for (i in seq_len(NROW(ds_ts_indicators))) {
  ts_indicator <- ds_ts_indicators$data[[i]]

  ts_indicator <- modify_indicator(
    ts_indicator = ts_indicator,
    modify_fun = new_attr_indcd,
    replace_exist = FALSE,
    date_index_field = "date",
    key_fields = "stkcd",
    parallel = parallel
  )
}

## End(Not run)

chriszheng2016/zstmodelr documentation built on June 13, 2021, 8:59 p.m.