# geom_text(aes(label=stars, color=value), size=8) + scale_colour_gradient(low="gre圓0", high="white", guide="none") + P <- ggplot(aes( x = country, y = variable, fill = value), data = plot.data)įig12 <- p + geom_tile() + scale_fill_gradient2( low = "#D7191C ", mid = "white ", high = "#2C7BB6 ") + Plot.data $ country <- factor( plot.data $ country, levels =levels(with( plot.data,], reorder( country, - value)))) # Sort country by highest by most significant coefficient Plot.data $ variable <- factor( plot.data $ variable, levels =levels(with( plot.data, reorder( variable, - value)))) # Sort coefficients by value
Plot.data $ p.value <- pnorm(abs( plot.data $ value), lower.tail = FALSE) * 2 # Calculate p-values for t-values Levels( plot.data $ variable) <- # Make coefficient names pretty Plot.data <- melt( values, id = "country ") # Convert to long Values $ country <- c( "All countries ", levels( barometer $ country.name)) # Add column with country names Values <- values # Remove taus only include actual coefficients Num.coefs <- length(attr(terms( form.expanded), "term.labels ")) # number of lhs coefficients Values <- ame(rbind(summary( ) $ coefficients,įUN = function ( x) summary( ]) $ coefficients)) # Create dataframe of t-values for combined model and all individual models # Figure 12: Grid plot for coefficient t-values (expanded)
0 kommentar(er)
