snippet2_blogpost

gistfile1.txt

# SHAP values: http://docs.h2o.ai/h2o/latest-stable/h2o-r/docs/reference/predict_contributions.H2OModel.html

SHAP_values <- predict_contributions.H2OModel(aml_leader, df_frame_split[[2]])

# Wrangling inspired here: https://bradleyboehmke.github.io/HOML/iml.html

shap_df <- SHAP_values %>%
  as.data.frame() %>%
  select(-BiasTerm) %>%
  gather(feature, shap_value) %>%
  group_by(feature) %>%
  mutate(shap_importance = mean(abs(shap_value)),
         shap_force = mean(shap_value)) %>% 
  ungroup()

# SHAP contribution plot
p1 <- ggplot(shap_df, aes(x = shap_value, y = reorder(feature, shap_importance))) +
  ggbeeswarm::geom_quasirandom(groupOnX = FALSE, varwidth = TRUE, size = 0.9, alpha = 0.5, width = 0.15) +
  xlab("SHAP value") +
  ylab(NULL) +
  theme_minimal(base_size = 15)

# SHAP importance plot
p2 <- shap_df %>% 
  select(feature, shap_importance) %>%
  distinct() %>% 
  ggplot(aes(x = reorder(feature, shap_importance), 
             y = shap_importance)) +
  geom_col(fill = 'black') +
  coord_flip() +
  xlab(NULL) +
  ylab("mean(|SHAP value|)") +
  theme_minimal(base_size = 15)

# Combine plots
gridExtra::grid.arrange(p1, p2, nrow = 1)