Hi team, We'd like to add a parameter `ymax` to `stats.boxcox_normmax`. This parameter will not only control the magnitude of the transformed data (addressing the overflow issue), but also give user the option to compute the unconstrained optimal lambda for Box-Cox transformation.
The default value of `ymax` is the maximum value of the input dtype, ensuring the transformed data does not cause overflow. If `ymax=np.inf`, users can compute the unconstrained optimal lambda. Other positive values of `ymax` are also accepted for more flexibility. Feel free to share your thoughts or any suggestions. https://github.com/scipy/scipy/pull/19631 Best, Xuefeng _______________________________________________ SciPy-Dev mailing list -- [email protected] To unsubscribe send an email to [email protected] https://mail.python.org/mailman3/lists/scipy-dev.python.org/ Member address: [email protected]
