preprocess_jump_data: Preprocess LM jump test result

In YalDan/hf.econometrics: High Frequency Econometrics

Preprocess LM jump test result

Description

Convenience functionalities for subsetting the jump results, changing the significance level, splitting the jumps into different waves, adding a Bonferroni correction, splitting the data by exchange / symbol, and modifying the variation measure for the test statistic

Usage

preprocess_jump_data(
  DATA,
  subs_date = NA,
  sign_level = 0.01,
  split_waves = FALSE,
  wave_dist = NA,
  wave_size = 5,
  bonferroni = TRUE,
  DIFF_ID = FALSE,
  asymptotic_variation = FALSE,
  hourly = FALSE
)

Arguments

DATA	The output from LM_JumpTest_2012.
subs_date	A vector with dates can be provided to subset from. Defaults to NA.
sign_level	The significance level for detecting jumps. Defaults to 0.01.
split_waves	Should the data be split into jump waves? Defaults to FALSE.
wave_dist	If split_waves = TRUE, where to set the splits? E.g. if no jump for longer than wave_dist = 5, the wave is over. This is adaptive w.r.t. to minutely / secondly frequency. Defaults to NA.
wave_size	Lower threshold of jumps needed to be in a wave for it to be considered? Defaults to 5, i.e. only waves with at least 5 jumps will be kept in the end. Set to 1 for all jumps to be included.
bonferroni	Should a Bonferroni correction be performed? Defaults to TRUE.
DIFF_ID	Should the data be evaluated by exchange and symbol or just by exchange? Defaults to FALSE, i.e. jumps are calculated per symbol.
asymptotic_variation	Should the asymptotic variation be used for calculating the test statistic? Defaults to FALSE, s.t. the empirical variation is used.
hourly	Is the test statistic calculated hourly? This can be useful for very high frequency data. Defaults to FALSE.

YalDan/hf.econometrics documentation built on May 10, 2024, 2:18 a.m.