Version: 3.19

mmm_attribution

R

library(ChannelAttributionPro)

token <- "yourtoken"

Data <- read.csv("https://app.channelattribution.io/data/data_mmm_2.csv", sep = ",")

D_variables <- data.frame(
  variable = c(
    "direct_searches",
    "facebook_impressions",
    "facebook_clicks",
    "google_impressions",
    "google_clicks"
  ),
  channel = c(
    "direct",
    "facebook_ads",
    "facebook_ads",
    "google_ads",
    "google_ads"
  ),
  measure = c(
    "direct_searches",
    "impressions",
    "clicks",
    "impressions",
    "clicks"
  ),
  stringsAsFactors = FALSE
)

target <- "conversions"

res <- mmm_attribution(
  Data = Data,
  D_variables = D_variables,
  target = target,
  model = "reward",
  model_score_target = "delta",
  max_p = 12,
  seed = 1234567,
  verbose = 1,
  password = token
)

print(head(res))

Using a different internal model:

res_copula <- mmm_attribution(
  Data = Data,
  D_variables = D_variables,
  target = target,
  model = "copula",
  model_score_target = "level",
  max_p = 12,
  nsim = 1000,
  seed = 1234567,
  password = token
)

res_linear <- mmm_attribution(
  Data = Data,
  D_variables = D_variables,
  target = target,
  model = "linear",
  model_score_target = "level",
  max_p = 12,
  password = token
)

To inspect diagnostics:

diag <- mmm_attribution(
  Data = Data,
  D_variables = D_variables,
  target = target,
  model = "reward",
  model_score_target = "delta",
  max_p = 12,
  return_diagnostics = TRUE,
  password = token
)

attribution <- diag$attribution
print(head(attribution))
print(diag$input_saturation_alpha_by_measure)
print(diag$model)

R​

R