In this section, we conduct a brief exploratory data analysis, visualizing the overall distributions of the chemical abundance data, the galactic coordinates of the stars, and pairwise relationships between the chemical abundance features. For simplicity, all plots shown here are based on the mean-imputed training data.
# get X (using mean-imputed) and metadata for EDA
metadata <- metadata$train
feature_modes <- list("small" = 7, "medium" = 11, "big" = 19)
train_data_ls <- purrr::map(
feature_modes,
~ get_abundance_data(data_mean_imputed$train, feature_mode = .x)
)
features_ls <- purrr::map(train_data_ls, ~ colnames(.x))
X <- train_data_ls$big# plot overall distribution
ggwrappers::plot_histogram(X) +
ggplot2::facet_wrap(~ variable, scales = "free_x") +
ggplot2::labs(x = "Data")
# plot boxplots per GC
plt_df <- dplyr::bind_cols(
X,
metadata |> dplyr::select(GC_NAME)
) |>
dplyr::group_by(GC_NAME) |>
dplyr::mutate(
GC_NAME = sprintf("%s (n = %d)", GC_NAME, dplyr::n())
) |>
dplyr::ungroup()
plt_vars <- sort(colnames(X))
plt_ls <- list()
for (plt_var in plt_vars) {
plt_ls[[plt_var]] <- plt_df |>
ggplot2::ggplot() +
ggplot2::aes(
x = reorder(GC_NAME, !!rlang::sym(plt_var)),
y = !!rlang::sym(plt_var),
fill = GC_NAME
) +
ggplot2::geom_boxplot() +
ggplot2::labs(x = "GC Name") +
vthemes::theme_vmodern() +
ggplot2::theme(
axis.text.x = ggplot2::element_text(
angle = 90, hjust = 1, vjust = 0.5
),
legend.position = "none"
)
}
plt <- patchwork::wrap_plots(plt_ls, ncol = 2) +
patchwork::plot_layout(axis_titles = "collect")
subchunkify(
plt, fig_height = 30, fig_width = 10,
caption = "'Distribution of abundance values per feature and GC in (mean-imputed) training data.'"
)
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by GC_NAME.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by GC_NAME.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by AL_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by AL_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by C_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by C_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by CA_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by CA_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by CI_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by CI_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by CO_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by CO_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by CR_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by CR_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by FE_H.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by FE_H.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by K_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by K_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by MG_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by MG_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by MN_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by MN_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by N_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by N_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by NA_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by NA_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by NI_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by NI_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by O_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by O_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by S_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by S_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by SI_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by SI_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by TI_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by TI_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by TIII_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by TIII_FE.
Unjittered scatter plot of the star locations (given by galactic coordinates), colored by V_FE.
Jittered scatter plot of the star locations (given by galactic coordinates), colored by V_FE.
