You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Since there exist DEC operators that depend on the dimension of the space they are applied on, such as the Lie Derivative and the Interior Product, we should organize these dimension specific implementations by letting the user choose separate default_dec_generate versions for 1d and 2d. Here's an example for 1d:
functiondefault_dec_generate_1d(...)
op =@match my_symbol begin:L=>lie_derivative_1d(...)
:i=>interior_product_1d(...)
...# For any operators that don't depend on dimensions like d, or hodge star
_ =>default_dec_generate(...)
end...
While we could do the same by passing a dimension parameter into the current default_dec_generate and then having the Lie Derivative and Interior Product implementations accept that dimension information, I feel this method organizes the operators a bit better for the user.
The text was updated successfully, but these errors were encountered:
GeorgeR227
changed the title
Seperate default_dec_generate versions for 1d and 2d.
Seperate default_dec_generate versions for 1d and 2d
Sep 7, 2023
Yes, we should also rethink whether the current method of passing hodge as a named argument to generate is scalable. Since there are also multiple sharps now, for example.
Since there exist DEC operators that depend on the dimension of the space they are applied on, such as the Lie Derivative and the Interior Product, we should organize these dimension specific implementations by letting the user choose separate default_dec_generate versions for 1d and 2d. Here's an example for 1d:
While we could do the same by passing a dimension parameter into the current default_dec_generate and then having the Lie Derivative and Interior Product implementations accept that dimension information, I feel this method organizes the operators a bit better for the user.
The text was updated successfully, but these errors were encountered: