Two stacked panels on a shared axis of adstocked spend, the saturation
curve's input (see response_curve()): the response and its marginal
return. With observed, levels below or above its range are shaded,
because there the curve is extrapolation. The point where the marginal
return is largest in size is marked as the steepest, unless it is the first
or last point.
Usage
# S3 method for class 'mm_response_curve'
plot(x, observed = NULL, ...)Arguments
- x
A data frame from
response_curve().- observed
Optional numeric vector of the levels seen in the data, normally the channel's adstocked series; its range is left unshaded.
- ...
Graphical arguments:
mainandylabreplace the upper panel's title and axis label,xlabthe spend axis label, andcolthe line colour. Others are ignored.
Examples
data(mm_weekly)
north <- mm_weekly[mm_weekly$geo == "north", ]
rc <- response_curve(seq(0, 6000, by = 100), coefficient = 6000,
half_max = 2250, shape = 1.6)
# Shade the levels outside the range of television's adstocked series
plot(rc, observed = adstock_geometric(north$tv, decay = 0.85))