DOC: multilevel tick example #27411
DOC: multilevel tick example #27411
Conversation
e52f7d4 to
c39134b
Compare
| # label the months: | ||
| sec = ax.secondary_xaxis(location=-0.075) | ||
| sec.xaxis.set_major_locator(mdates.MonthLocator(bymonthday=1)) | ||
| sec.xaxis.set_major_formatter(mdates.DateFormatter(' %b')) |
There was a problem hiding this comment.
I assume you add the spaces before the %b to improve the alignment with the tick. I just spent longer than I care to admit trying to figure out a way to set the horizontal alignment of the labels, and now I think I understand why you want to overhaul Axis!
For the record, this works, but I'm not sure it would be an improvement.
fig.draw_without_rendering() # Otherwise only one label exists to loop through. Is there a cheaper way to force tick generation?
for label in sec.xaxis.get_ticklabels():
label.set_horizontalalignment('left')Is it worth a comment to explain the spaces though?
There was a problem hiding this comment.
Yes, it's all pretty annoying. And of course that alignment only holds for ticks that presently exist - if you pan off to right or left the new ticks won't get the alignment.
There was a problem hiding this comment.
aI think lighter lift to improve this would be to add alignment to tick_params #20644
| # This example also differs from the above, in that we placed it at a | ||
| # location below the main axes ``location=-0.075`` and then we hide the spine | ||
| # by setting the line width to zero. That means that our formatter no longer | ||
| # needs the carriage returns of the previous two examples. |
There was a problem hiding this comment.
Maybe add a sentence near the top somewhere to explain why we do add the "\n" in those examples.
|
|
||
| ax.plot(['cats', 'dogs', 'pigs', 'snakes', 'lizards', 'chickens', | ||
| 'eagles', 'herons', 'buzzards'], | ||
| rng.normal(size=9)) |
There was a problem hiding this comment.
| rng.normal(size=9)) | |
| rng.normal(size=9), 'o') |
I know it doesn't matter a jot for the purpose of the example, but lines between categorical data don't make sense to me.
| sec.set_xticks([5, 15, 25], labels=['\nOughts', '\nTeens', '\nTwenties']) | ||
|
|
||
| # %% | ||
| # This example adds a second level of annotation to a categorical axis. Note |
There was a problem hiding this comment.
| # This example adds a second level of annotation to a categorical axis. Note | |
| # The second example adds a second level of annotation to a categorical axis. Note |
To make it really obvious to the reader that we are moving on.
| # This example adds a second level of annotation to a categorical axis. Note | ||
| # that here we need to note that each animal (category) is assigned an integer, |
There was a problem hiding this comment.
| # This example adds a second level of annotation to a categorical axis. Note | |
| # that here we need to note that each animal (category) is assigned an integer, | |
| # This example adds a second level of annotation to a categorical axis. | |
| # Here we need to note that each animal (category) is assigned an integer, |
Scans a little better for me.
|
|
||
| # %% | ||
| # Dates are another common place where we may want to have a second level of tick | ||
| # labels. Here we take advantage of the ability to add an automatic locator and |
There was a problem hiding this comment.
| # labels. Here we take advantage of the ability to add an automatic locator and | |
| # labels. Next we take advantage of the ability to add an automatic locator and |
Moving on again.
c39134b to
8405388
Compare
doc/devel/index.rst
Outdated
| `development version <https://matplotlib.org/devdocs/devel/index.html>`_ | ||
| of this document as it will have the most up to date installation | ||
| instructions, workflow process, and contributing guidelines. | ||
|
|
There was a problem hiding this comment.
Where did this come from? I didn't see it before and it doesn't show in the "force-push" diff 😕
There was a problem hiding this comment.
I've had that happen for this PR a couple of times where changes on main get shown as diffs in this PR? Not sure why, but a rebase on main fixes it
8405388 to
f962e7a
Compare
fix typo Co-authored-by: Ruth Comer <10599679+rcomer@users.noreply.github.com>
|
@meeseeksdev backport to v3.8.x |
|
@meeseeksdev backport to v3.8.2-doc |
…411-on-v3.8.2-doc Backport PR #27411 on branch v3.8.2-doc (DOC: multilevel tick example)
…411-on-v3.8.x Backport PR #27411 on branch v3.8.x (DOC: multilevel tick example)
This adds an example showing one method of adding multilevel ticks to a plot.
Not a substitute for #6321, but shows that it's possible with minimal fussing. Also partially addresses #27394, or at least gives an example of how to achieve the date axis effect discussed there