MultiCellWinner: Multi-Cell Winner Declaration Method for Market Selection.

View source: R/MultiCell.R

MultiCellWinnerR Documentation

Multi-Cell Winner Declaration Method for Market Selection.

Description

[Experimental]

MultiCellWinner explores the likelihood of observe a winning cell for a given Multi-Cell Market Selection. This method analyzes all pairwise comparisons to determine how much larger the incremental ROAS (iROAS) must be for a cell to be declared the winner based on a statistical significance test.

Usage

MultiCellWinner(
  multicell_power_obj,
  effect_size = NULL,
  geolift_type = "standard",
  ROAS = seq(0, 5, 0.05),
  alpha = 0.1,
  method = "conformal",
  stat_test = "Total"
)

## S3 method for class 'MultiCellWinner'
print(x, ...)

Arguments

multicell_power_obj

A MultiCellPower object.

effect_size

A numeric value representing the lift to be simulated across all cells. If not specified (default), the algorithm will use the largest lift needed to obtain a well-powered test across all cells.

geolift_type

String that specifies the type of GeoLift test to be performed:

  • "standard": Standard GeoLift test where ads are shown to test regions. Defualt.

  • "inverse": Inverse or Negative GeoLift test where the test group is holded-out from the treatment.

ROAS

Vector of incremental Return on Ad Spend (iROAS) values to assess. Set to seq(0,5,0.05) by default.

alpha

Significance Level. By default 0.1.

method

A string indicating the method used to calculate the aggregate ATT Confidence Intervals.

  • "conformal": Conformal Inference. Defualt.

  • "jackknife+": Jackknife+ (exclusively for stat_test = "Total").

stat_test

A string indicating the test statistic.

  • "Total": The test statistic is the sum of all treatment effects, i.e. sum(abs(x)). Default.

  • "Negative": One-sided test against positive effects i.e. -sum(x). Recommended for Negative Lift tests.

  • "Positive": One-sided test against negative effects i.e. sum(x). Recommended for Positive Lift tests.

x

MultiCellWinner()

...

Optional arguments

Value

A list with two objects:

  • "results": Data frame with all pairwise comparisons and required iROAS needed to declare a winner for the multi-cell test.

  • "simulations": The complete data frame of all simulations for all pairwise comparisons.


facebookincubator/GeoLift documentation built on May 31, 2024, 10:09 a.m.