Report timings.generic pandas.data frame.to type interface#7205
Conversation
* Port onto latest pandas and matplotlib (bump pandas dep). * Remove back-compat code. * Add basic tests.
"to_<output>" methods. Added "json" and "csv" to the list of raw output formats. n.b: You can't use `to_excel` or `to_markdown` without additional optional dependencies. Also left scope in the interface for "csv-specific-char-delimited".
wxtim
force-pushed
the
report-timings.generic-pandas.DataFrame.to_type_interface
branch
from
ffef099 to
6e04161
Compare
| """ | ||
|
|
||
| buf.write('<html><head><style>%s</style></head><body>' % css) | ||
| buf.write('<!DOCTYPE html><html><head><style>%s</style></head><body>' % css) |
There was a problem hiding this comment.
Not having this showed as an warning when I looked at the HTML with dev tools on
| 'raw': format_raw, | ||
| 'summary': format_summary, | ||
| 'html': format_html, | ||
| 'csv': partial(format_generic, format='csv'), |
There was a problem hiding this comment.
Sadly to_markdown and to_excel require extra dependencies, and to_html, though it works isn't very nice to look at and needs some CSS appended.
|
Might as well go with the CSV and JSON outputs added here, it's adding no more to dependencies than what was already there. Some Context: I was kinda hoping that we might rewrite the data handling side of this code to remove that However, it does look like these dependencies are delivering real value here (especially the HTML output) and there may be value to this post-analysis view, even though we probs don't want to take on the burden of maintaining both. |
2 participants
Follow on from #7204
Should satisfy the desires of any users who wish to open up the Raw data in Excel.
Check List
CONTRIBUTING.mdand added my name as a Code Contributor.setup.cfg(andconda-environment.ymlif present).?.?.xbranch.