Top educational background for LMO2023 skill based clusters

Source code:

Intro

In the 2023 LMO we formed clusters of occupations on the basis of skill similarity. This provided two useful bits of information:

1. What occupations share similar skill profiles: useful information for those considering a career change.
2. What are the relative skill differences that distinguish the clusters: useful information for those looking for an occupation that requires their relative skill strengths.

Here we take a step back and identify what educational backgrounds are most common for people working in a given cluster. This is useful information for those who know in what cluster they wish to work, but need to choose an educational path that leads to that cluster.

Code

temp <- inner_join(edu_noc, skill_clusters, by=join_by("NOC"))|>
  group_by(new_cluster, Education)|>
  summarize(value=sum(value))|>
  mutate(value=value/sum(value))|>
  group_by(new_cluster, .add=FALSE)|>
  slice_max(value, n=30)|>
  mutate(Education2=Education)|>
  separate(Education2, into = c("CIP","Highest Attainment"), sep=": ")|>
  mutate(Education=tidytext::reorder_within(Education, value, new_cluster))

plt <- ggplot(temp, aes(value, 
                        Education,
                        fill=`Highest Attainment`,
                        text=CIP))+
  geom_col()+
  scale_x_continuous(labels=scales::percent)+
  scale_y_discrete( expand = c(0.1, 0.1))+
  facet_wrap(~new_cluster, scales = "free_y")+
  theme_minimal()+
  labs(x=NULL, y=NULL)+
  theme(axis.text.y=element_blank(), 
        axis.ticks.y=element_blank(),
        axis.text.x = element_text(angle = 45, vjust = 1, hjust=1)
        )

ggplotly(plt, tooltip = "text")|>
  config(displayModeBar = FALSE)|>
  layout(legend = list(
      orientation = "h"
    )
  )