ENH: Adding DataFrame plotting benchmarks for large datasets #61546
Conversation
|
@rhshadrach do you have an opinion about this? |
asv_bench/benchmarks/plotting.py
Outdated
|
|
||
| def time_plot_large_dataframe_single_column(self, size, datetime_index): | ||
| """Baseline: plotting single column for comparison""" | ||
| self.df.iloc[:, 0].plot() |
There was a problem hiding this comment.
Can you do the iloc operation in setup so its not part of the timing?
shadnikn
force-pushed
the
ENH-61532-speedup-dataframe-plot
branch
from
54ee8b8 to
2324563
Compare
|
@shadnikn - what is the runtime of these benchmarks on your machine? |
| [(1000, 10), (1000, 50), (1000, 100), (5000, 20), (10000, 10)], | ||
| [True, False] # DatetimeIndex or not |
There was a problem hiding this comment.
I think this is too many cases. Can you just do [(1000, 100)]
|
|
||
|
|
||
| from .pandas_vb_common import setup # noqa: F401 isort:skip | ||
| class DataFramePlottingLarge: |
There was a problem hiding this comment.
I think what differentiates this from FramePlotting is that these are wide. Can you rename this class to FramePlottingWide.
| [(1000, 10), (1000, 50), (1000, 100), (5000, 20), (10000, 10)], | ||
| [True, False] # DatetimeIndex or not | ||
| ] | ||
| param_names = ["size", "datetime_index"] |
There was a problem hiding this comment.
Can you rename size -> shape.
|
Thanks for the pull request, but it appears to have gone stale. If interested in continuing, please merge in the main branch, address any review comments and/or failing tests, and we can reopen. |
4 participants
DataFrame.plotusingLineCollection#61532 - Adding in performance benchmarks for DataFrame plotting with large datasets.DataFrame.plotwithDatetimeIndex#61398 and ENH: speed upDataFrame.plotusingLineCollection#61532. Tests multiple DataFrame sizes with/w/o DatetimeIndex & provides a baseline single-column comparison.- DataFrame sizes: (1000,10) to (10000,10)
- DatetimeIndex vs. regular index comparison
- Multi-column vs. single-column plotting.