todo.TODO

Julia:
  ☐ Interpolation Routines
    ✔ Standard @done (13-12-06 11:20)
    ☐ Derivatives type 1 @started(13-12-06 11:20)
    ☐ Derivatives type 2
    ☐ Derivatives type 3
  ✔ Anisotropic grids @done (13-11-27 23:51)
  ✔ Non [-1, 1]^d grids/routines (see dolo) @done (13-12-06 11:20)
  ☐ Check for combinations with replacement function in base (or statsbase) @spencer 06-21-2014 01:36 AM
  Ideas:
    ☐ I noticed that for (40, 2) there are 10764961 points in B, but only 3090461 are non-zero. That is only 28.7085%. Maybe we should use sparse matrices. We should do some testing to see if that would help. @spencer @started(13-11-08 01:34)
  Speed Improvements:
   ☐Try to use the inbounds macro to speed up indexing -- this eliminates bounds checking @spencer 2013-11-21 10:26
   ✔ This is big re-write, but write things in column-major form. This should indexing quite a bit more efficient @done (13-11-29 14:18)
   ✔ Add type checks in the middle of functions saves *A ton* of time. Trying to use build_B(15, 3) from Steven without them ran for over a minute before I stopped it. With the type checks it runs in 0.33 seconds. (2013-11-29 14:20) @done (13-12-06 11:20)

  Style:
    ☐ I removed all the docstrings. I should add some comments somewhere. @spencer 06-21-2014 01:35 AM


Python:
 ✔ Interpolation Routines @done (13-12-03 12:51)
 ✔ Anisotropic grids @done (13-11-27 23:51)
 ✔ Non [-1, 1]^d grids/routines (see dolo) @done (13-12-03 12:51)

General Improvements:
  ☐ Improve the build B function.  We spend 90%+ of our time stuck building it. @chase
    ☐ mu = 2 -- done in python
    ☐ mu = 3 -- close in python
    ☐ General mu -- haha


Docs:
  ☐ Make it so users can go next and previous from module pages (right now they have to go pack to the Smolyak Package page) 2013-12-03 12:52
  ☐ Add interpolation page 2013-12-06 11:20