def parameter_importance():
import parameter_importance as pi
# find datasets to fill in the selector
selections = utils.select_init_data(request.args.get("id", ""))
dataset = selections["current_dataset"]
# get parameters for this dataset
(cols, low_var_cols) = pi.get_params(dataset)
# rank parameters
(rank1, _, rank2, scores) = pi.rank(dataset)
# correlation analysis
(rank0, correlation_plots) = pi.correlation_analysis(dataset)
# plot
scatter_plot = pi.scatter_plot(dataset, request.args.get("varx", rank1[0]), request.args.get("vary", rank1[1]))
scores_plot = pi.rlasso_scores_plot(rank2, scores)
return render_template(
"parameter_importance.html",
cols=cols,
low_var_cols=low_var_cols.tolist(),
selections=selections,
scatter_plot=scatter_plot,
scores_plot=scores_plot,
correlation_plots=correlation_plots,
rank0=rank0[:9],
rank1=rank1[:9],
)
def initial_data():
import initial_data
# find datasets to fill in the selector
selections = utils.select_init_data(request.args.get("id", ""))
# do plots
(low_var_cols, var_percent, correlation_plot, pca_plot) = initial_data.produce_plots(selections["current_dataset"])
return render_template(
"initial_data.html",
selections=selections,
correlation_plot=correlation_plot,
low_var_cols=low_var_cols,
var_percent=var_percent,
pca_plot=pca_plot,
)
