Yahoo options failing again in the docs

[jorisvandenbossche]

See http://pandas-docs.github.io/pandas-docs-travis/remote_data.html#yahoo-finance-options and http://pandas-docs.github.io/pandas-docs-travis/whatsnew.html#whatsnew-0151-api

@dstephens99

[jorisvandenbossche]

Maybe related to #8783, but the examples in the docs were not yet failing when that PR was merged.
Isn't it also strange there are no tests failing when the examples in the docs fail?

[jreback]

these look ok
what is failing?

[jorisvandenbossche]

ah, these were failing a couple of hours ago. Maybe the yahoo page that was temporarily not available ..

[jreback]

@jorisvandenbossche I believe it!

[jorisvandenbossche]

reopening this for now, as they are failing again. We should just be sure there are no failures when building the docs for the 0.15.2 release

http://pandas-docs.github.io/pandas-docs-travis/remote_data.html#yahoo-finance-options

[jreback]

looks ok to me. These might fail outside of mkt hours.

[jorisvandenbossche]

@jreback ah, are there hours that this function always fails? That could be of course the reason that I regularly see failures.

[jreback]

I would think they would always are available, but its possible (this is a guess), that > 12AM EST - 3AM ST (e.g. middle of the night EST) that they might not be.

[davidastephens]

I'll check at some random times to see if I can isolate the error (maybe the page changes in some way). If you see it fail, can you post the traceback?

[jorisvandenbossche]

You can see the traceback in the logs on travis (eg https://travis-ci.org/pydata/pandas/jobs/43288475, scroll down and expand line 1216 ('$ if [ -f /tmp/doc.log ]; then cat /tmp/doc.log; fi')). There you see the doc build log.

[davidastephens]

Thanks. The tests aren't failing because we skip on remote data error - this is failing because its not receiving data from Yahoo.

Maybe its worth setting it up so that it retries 3 times when it doesn't get any data - that might reduce the number of times it errors out.

[jreback]

I think this is fixed. If its not pls reopen.