Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | ENH: missingdata: Have some basic assignment and indexing with NA working | Mark Wiebe | 2011-08-27 | 1 | -5/+11 |
| | |||||
* | ENH: missingdata: Add the NA mask members to PyArrayObject | Mark Wiebe | 2011-08-27 | 1 | -2/+15 |
| | |||||
* | STY: Updates for pull request feedback from Chuck and Ben | Mark Wiebe | 2011-07-20 | 1 | -9/+124 |
| | |||||
* | ENH: core: Deprecating direct access to the PyArrayObject fields | Mark Wiebe | 2011-07-19 | 1 | -8/+8 |
| | | | | NOTE: WIP, code doesn't build | ||||
* | NEP: missingdata: Rename 'namask' to 'maskna', which looks better | Mark Wiebe | 2011-07-19 | 1 | -30/+40 |
| | |||||
* | NEP: missingdata: Add a small note about single NA vs multi-NA | Mark Wiebe | 2011-07-08 | 1 | -0/+8 |
| | |||||
* | NEP: missingdata: Describe a possible way multi-NA could be added in the future | Mark Wiebe | 2011-07-08 | 1 | -0/+27 |
| | |||||
* | NEP: missingdata: Change the mask definition | Mark Wiebe | 2011-07-08 | 1 | -6/+17 |
| | | | | | | | This change allows npy_bool to be a mask which always has payload zero. This doesn't let combining masks with payloads to be a simple 'min' operation as the previous design, but allowing npy_bool as the mask appears to be a very worthwhile tradeoff. | ||||
* | NEP: missingdata: Design the mask binary format so it's extensible with a ↵ | Mark Wiebe | 2011-07-08 | 1 | -7/+21 |
| | | | | payload | ||||
* | ENH: nditer: Add the per-operand masked iteration flags | Mark Wiebe | 2011-07-08 | 1 | -0/+4 |
| | |||||
* | NEP: missingdata: Adjust the parenthetical part of the NA/IGNORE defn | Mark Wiebe | 2011-07-07 | 1 | -3/+3 |
| | |||||
* | NEP: missingdata: Add copyright info, some more NAMASK iterator clarification | Mark Wiebe | 2011-07-07 | 1 | -6/+17 |
| | |||||
* | NEP: missingdata: Add more C API details | Mark Wiebe | 2011-07-07 | 1 | -1/+27 |
| | |||||
* | NEP: missingdata: Start some details about masked iteration | Mark Wiebe | 2011-07-07 | 1 | -6/+49 |
| | |||||
* | NEP: missing-data: Add Peter to the Acknowledgments | Mark Wiebe | 2011-07-06 | 1 | -0/+1 |
| | |||||
* | NEP: missing-data: Add "Python API" and "C API" definitions | Mark Wiebe | 2011-07-06 | 1 | -1/+13 |
| | |||||
* | NEP: missing-data: Add numpy.ma to the glossary | Mark Wiebe | 2011-07-06 | 1 | -1/+5 |
| | |||||
* | NEP: missing-data: Add glossary of terms, try to clarify them better | Mark Wiebe | 2011-07-06 | 1 | -28/+59 |
| | | | | | This is a step towards having everyone on the list use the same vocabulary with specific nailed-down definitions for the terms. | ||||
* | NEP: missing-data: Fix copy/paste/edit typo for np.all example | Mark Wiebe | 2011-07-06 | 1 | -6/+6 |
| | |||||
* | NEP: missing-data: Incorporate Ben's feedback, add section on 'shared masks' | Mark Wiebe | 2011-07-06 | 1 | -65/+107 |
| | | | | This includes an email comment from Ben about 'np.any' and 'np.all'. | ||||
* | NEP: missingdata: Rename 'hasmask' and friends to 'hasnamask' and friends | Mark Wiebe | 2011-07-06 | 1 | -7/+46 |
| | | | | | This is an attempt to reduce potential confusion between the existing numpy.ma and the NA-supporting built-in mask. | ||||
* | NEP: missing-data: Add name to acknowledgments | Mark Wiebe | 2011-07-06 | 1 | -0/+1 |
| | |||||
* | NEP: missingdata: Various improvements | Mark Wiebe | 2011-07-06 | 1 | -20/+39 |
| | |||||
* | NEP: Trying to make the NEP's position on missing values and masks vs bit ↵ | Mark Wiebe | 2011-07-06 | 1 | -22/+84 |
| | | | | patterns more clear | ||||
* | ENH: umath: Add tests, work out kinks in ufunc 'where=' parameter | Mark Wiebe | 2011-07-06 | 1 | -0/+2 |
| | |||||
* | ENH: umath: Add parsing of a 'where=' parameter in the element-wise ufunc | Mark Wiebe | 2011-07-06 | 1 | -8/+10 |
| | |||||
* | ENH: missingdata: Move getting the masked ufunc inner loop to a different ↵ | Mark Wiebe | 2011-07-06 | 1 | -0/+1 |
| | | | | | | | | | function This is so that backwards compatibility can be retained with denying regular ufuncs from supporting dynamic auxiliary data, but allowing it for the masked inner loops so that the default masked loop can store the regular inner loop it proxies calls to. | ||||
* | ENH: missingdata: Add a mechanism to get masked versions of ufunc inner loops | Mark Wiebe | 2011-07-06 | 1 | -2/+10 |
| | |||||
* | DOC: A few typos in the missing data NEP. | Keith Goodman | 2011-06-27 | 1 | -3/+4 |
| | |||||
* | NEP: missing-data: Rename the NEP as it now describes more generality than ↵ | Mark Wiebe | 2011-06-27 | 1 | -0/+529 |
just masks |