setPyModel: Set python model

In Vitalic57/stratbuilder2pub: It Provides Tools to Create and Backtest Financial Models

Description

User can create realization of python class Model inherited from BaseModel. All dependencies user should specify in Dockerfile, then with help of addDocker push docker to gitlab.

Usage

setPyModel(this, pyfile, dockername, lookback_init, lookback_step = 0,
  lookback = 0, data = quote(spread), pathwise = FALSE, as = "signal",
  hostname = "localhost", modelpath = NULL, update_with_betas = FALSE,
  vector_step = FALSE, args = list(), port = 4000)

## S3 method for class 'modelStrategy'
setPyModel(this, pyfile, dockername, lookback_init,
  lookback_step = 0, lookback = 0, data = quote(spread),
  pathwise = FALSE, as = "signal", hostname = "localhost",
  modelpath = NULL, update_with_betas = FALSE, vector_step = FALSE,
  args = list(), port = 4000)

Arguments

this	modelStrategy
pyfile	character, path to .py file
dockername	character, name of docker container
lookback_init	numeric, periods of time for training the model
lookback_step	numeric, periods of time for step of the model
lookback	numeric, how many periods need to be to evaluate data expression
data	quote, expression that returns numeric vector or matrix
pathwise	logical, if TRUE then http requests will be send on each iteration
as	character, name of variable that will contains results of model evaluation
hostname	character, name of host.
modelpath	character, path to pretrained model
update_with_betas	logical, by default FALSE. If TRUE then model will be reinit after reinitilization of betas weights
vector_step	logical, if TRUE then model.py file should have ability to take vector to step and output vector after predict
args	list, list of arguments for python file
port	numeric, port for docker container

Vitalic57/stratbuilder2pub documentation built on Dec. 18, 2019, 2:56 a.m.