label_meta: Meta labeling, including three options: triple barriers,...

In larryleihua/fmlr: Financial Machine Learning using R

Description

Meta labeling, including three options: triple barriers, upper and vertical barriers, and lower and vertical barriers.

Usage

label_meta(x, events, ptSl, n_ex = 0)

Arguments

x	a vector of prices series to be labeled.
events	a dataframe that has the following columns: t0: event's starting time index. t1: event's ending time index; if t1==Inf then no vertical barrier, i.e., last observation in x is the vertical barrier. trgt: the unit absolute return used to set up the upper and lower barriers. side: 0: both upper and lower barriers; 1: only upper; -1: only lower.
ptSl	a vector of two multipliers for the upper and lower barriers, respectively.
n_ex	number of excluded observations at the begining of x; default is 0.

Value

data frame with the following columns:

• T_up: local time index when the upper barrier is hit; Inf means that upper is not hit.

• T_lo: local time index when the lower barrier is hit; Inf means that lower is not hit.

• t1: local time index when the vertical barrier is hit.

• ret: return associated with the event.

• label: low:-1, vertical:0, upper:1.

• t0Fea: begining time index of feature bars.

• t1Fea: ending time index of feature bars.

• tLabel: ending time index of events, i.e., when the labels are created. Both t1Fea and tLabel will be useful for sequential bootstrap.

Author(s)

Larry Lei Hua

larryleihua/fmlr documentation built on Jan. 31, 2020, 8:55 a.m.