FIX: revert part of #11770 #13131
FIX: revert part of #11770 #13131
Conversation
jorisvandenbossche
added
Bug
Datetime
| SEC_PER_DAY = SEC_PER_HOUR * HOURS_PER_DAY | ||
| SEC_PER_WEEK = SEC_PER_DAY * DAYS_PER_WEEK | ||
|
|
||
| MUSECONDS_PER_DAY = 1e6 * SEC_PER_DAY |
There was a problem hiding this comment.
@tacaswell Are all these constants needed? (in the original PR, it were only 4 that you removed)
|
@tacaswell any tests? can you add a whatsnew note? (can be something as simple as compat for .....issue because of incorrect tick display with mpl < 2.0....) |
tacaswell
force-pushed
the
fix_autodate
branch
2 times, most recently
from
c3d8f27
to
d4df66b
Compare
Reverts part of: pandas-dev#11770 Closes: matplotlib/matplotlib#6365 The mistake in pandas-dev#11770 was missing that pandas had a 1/us not 1/s scaled bucket.
|
Only took 4 force-pushes to get there. @jorisvandenbossche Added back the minimal set of constants to compute what was needed. What sort of test do you want for this? You merged the last one with out a test 👿 |
|
can u create something that failed on the previous and now works? just maybe a check on axis ticks or something simple |
|
thanks @tacaswell pls verify with mpl >= 2.0 when you have a chance. |
|
did we have an issue where this came up on pnadas side? or was only on mpl side? |
|
Only on the MPL tracker. |
|
Sorry i took to long to get a test in. On Fri, May 20, 2016, 10:37 Tom Augspurger notifications@github.com wrote:
|
|
np ideally if u can post a couple of assertions we can add something to prevente future regressions |
Reverts part of:
#11770
Closes:
matplotlib/matplotlib#6365
The mistake in #11770 was missing that pandas had a 1/us not 1/s
scaled bucket.