I am currently using your fabulous interactions package to visualise some statistical models I am running. I especially appreciate the interoperability with ggplot2. However, one thing I cannot work out how to do is change point shape by a variable that is not involved in the interaction. Specifically, I have a categorical variable with three levels in my model (as a random factor), and I would like the points to be shaped according to the levels of that factor. I'm sure this is quite a basic question, but I cannot work out how to use the point.shape argument to map to a factor not involved in the interaction, or to otherwise add a shape aesthetic to the graph. Any advice you can give would be hugely appreciated.
EXAMPLE:
My graph as it currently stands has the following code:
graph<- interact_plot(model1, pred="Mean_AlphaPhy",
modx = "Disturbance", plot.points=TRUE, partial.residuals = TRUE) +
scale_y_continuous(trans='log10') +
xlab("Mean Alpha Phylogenetic Diversity") + ylab("Number of plants colonising") +
scale_color_manual(values=c("#999999", "#E69F00")) +
scale_linetype_manual(values=c("solid", "solid")) +
theme_bw() + guides(linetype = "none") + theme(legend.position = "top") +
theme(axis.text=element_text(size=7),
axis.title=element_text(size=9),
legend.text=element_text(size=7),
legend.title=element_text(size=9))
And it is based on this model:
model1 <- glmmTMB(germ_num ~ Mean_AlphaPhy*Disturbance + propagule_pressure +
suitabilitypc1V2abs + suitabilitypc2V2abs + Density +
(1|Family) + (1|POWO.name) + (1|Site) + (1|Study),
family=nbinom2(link = "log"),
data=germinationcount)
I would like the graph to be exactly as it is, but with the points having one of three shapes, determined by the levels of 'Study'.
I am currently using your fabulous interactions package to visualise some statistical models I am running. I especially appreciate the interoperability with ggplot2. However, one thing I cannot work out how to do is change point shape by a variable that is not involved in the interaction. Specifically, I have a categorical variable with three levels in my model (as a random factor), and I would like the points to be shaped according to the levels of that factor. I'm sure this is quite a basic question, but I cannot work out how to use the point.shape argument to map to a factor not involved in the interaction, or to otherwise add a shape aesthetic to the graph. Any advice you can give would be hugely appreciated.
EXAMPLE:
My graph as it currently stands has the following code:
graph<- interact_plot(model1, pred="Mean_AlphaPhy",
modx = "Disturbance", plot.points=TRUE, partial.residuals = TRUE) +
scale_y_continuous(trans='log10') +
xlab("Mean Alpha Phylogenetic Diversity") + ylab("Number of plants colonising") +
scale_color_manual(values=c("#999999", "#E69F00")) +
scale_linetype_manual(values=c("solid", "solid")) +
theme_bw() + guides(linetype = "none") + theme(legend.position = "top") +
theme(axis.text=element_text(size=7),
axis.title=element_text(size=9),
legend.text=element_text(size=7),
legend.title=element_text(size=9))
And it is based on this model:
model1 <- glmmTMB(germ_num ~ Mean_AlphaPhy*Disturbance + propagule_pressure +
suitabilitypc1V2abs + suitabilitypc2V2abs + Density +
(1|Family) + (1|POWO.name) + (1|Site) + (1|Study),
family=nbinom2(link = "log"),
data=germinationcount)
I would like the graph to be exactly as it is, but with the points having one of three shapes, determined by the levels of 'Study'.