warning("spline basis weights close to boundaries; try reducing spline_df")

[paulinetrinh]

suppressWarnings() about spline basis weights close to boundaries. Theoretically, having spline basis weights close to boundaries could change the asymptotics of our estimators. However, in our simulation output we didn't see this occurring. Currently, these warnings are being suppressed since there is no empirical evidence that it makes a difference in practice.

TODO: evaluate the impact of spline basis weights being close to boundaries on our estimators via simulation. If there is an issue consider using a bootstrap to address the problem

[adw96]

In the meantime (though I do want this labelled high priority), can we suppress the warnings at intermediate steps, and only output a warning if it happens on the final iteration before convergence?

[paulinetrinh]

Comparison of happi 2 vs. 3 knot splines on prausnitzii dataset for gene L-rhamnose isomerase

[paulinetrinh]

In the meantime (though I do want this labelled high priority), can we suppress the warnings at intermediate steps, and only output a warning if it happens on the final iteration before convergence?

Current implementation of happi suppresses warnings at intermediate steps and only outputs a warning if it happens on the final iteration