multi_term_arima: multi_term_arima

In tlcaputi/gtrendR: An R Complement to the gtrends Python package

multi_term_arima

Description

multi_term_arima

Usage

multi_term_arima(
  input_dir = "C:/Users/tcapu/Google Drive/modules/gtrendR/READMEcode/input",
  geo = "US",
  terms_to_use = NA,
  timeframe_to_use = NA,
  beginperiod = T,
  preperiod = 90,
  endperiod = T,
  interrupt = "2020-03-01",
  bootstrap = T,
  bootnum = 1000,
  na = "kalman",
  kalman = F,
  include_data = T,
  linear = F,
  min0 = F,
  logit = T,
  scale = T,
  alpha = 0.05,
  arima.approx = T
)

Arguments

beginperiod	How far back you want the "pre" period to go
preperiod	This creates a beginperiod but with a number of days instead of a date
endperiod	How far after the interruption you want to go
interrupt	The date where things change. ARIMA will be predicted on all days before the interrupt.
df	A dataframe including time as timestamp and searches for your given geography in one column.
scaletitle	Title of the scale
scalelimits	vector of two values for min and max for the scale
linecol	Line color
lowcol	Color for low values
midcol	Color for mid values
highcol	Color for high values
save	Default is True, If False, don't save
width	Width of file in inches
height	Height of file in inches
outfn	Output filename

Examples

multiterms <- multi_term_arima(

  ## A folder containing all of your gtrends data
  input_dir = "./input",

  ## Which data to use
  geo = "US", # Geography you want to use
  terms_to_use = NA, # Terms you'd like to analyze. If NA then all terms
  timeframe_to_use = NA, # Only analyze data with filenames that contain a certain timeframe. If NA then all timeframes


  ## Parameters of time periods
  beginperiod = T, # Beginning of the before period, if T then beginning of data
  preperiod = 90, # If beginperiod is logical, preperiod is the number of days before interrupt to include in before period
  endperiod = T, # End of the end period, if T then end of data
  interrupt = "2020-03-01", # Date for interruption, splitting before and after periods


  ## Analytical arguments
  bootstrap = T, # Bootstrap CIs
  bootnum = 1000, # Number of bootstraps
  kalman = T # If T, impute with Kalman
)

tlcaputi/gtrendR documentation built on Nov. 3, 2022, 10:46 p.m.