Plots weighted scatterplots for meta-analytic data. Can plot effect size as
a function of either continuous (numeric, integer) or categorical (factor,
character) predictors.

WeightedScatter(
data,
yi = "yi",
vi = "vi",
vars = NULL,
tau2 = NULL,
summarize = TRUE
)

## Arguments

data |
A data.frame. |

yi |
Character. The name of the column in `data` that contains the
meta-analysis effect sizes. Defaults to `"yi"` . |

vi |
Character. The name of the column in the `data` that contains
the variances of the effect sizes. Defaults to `"vi"` . By default,
`vi` is used to calculate fixed-effects weights, because fixed effects
weights summarize the data set at hand, rather than generalizing to the
population. |

vars |
Character vector containing the names of specific moderator
variables to plot. When set to `NULL` , the default, all moderators
are plotted. |

tau2 |
Numeric. Provide an optional value for tau2. If this value is
provided, random-effects weights will be used instead of fixed-effects
weights. |

summarize |
Logical. Should summary stats be displayed? Defaults to
FALSE. If TRUE, a smooth trend line is displayed for continuous variables,
using [stats::loess()] for less than 1000 observations, and [mgcv::gam()] for
larger datasets. For categorical variables, box-and-whiskers plots are
displayed. Outliers are omitted, because the raw data fulfill this function. |

## Value

A gtable object.

## Examples

#> Warning: Computation failed in `stat_boxplot()`:
#> package 'SparseM' was installed before R 4.0.0: please re-install it

#> Warning: Computation failed in `stat_boxplot()`:
#> package 'SparseM' was installed before R 4.0.0: please re-install it