Fix iplot offline #599

Kully · 2016-11-01T14:50:47Z

Re iplot issue brought up in #593 (comment)

Fixed by just removing the unnecessary show_link and link_text from the _plot_html() call in this line.

plot() looks fine as it has show_link and link_text in the signature.

Kully · 2016-11-01T15:12:53Z

Tests passed locally but getting strange encoding error on circle. It looks like its in site-packages which I think is core Python. Any thoughts @theengineear ?

cc: @choldgraf

Kully · 2016-11-01T16:01:38Z

The last bit of the error message:

Cleaning up...
Command /home/ubuntu/.pyenv/versions/2.7.8/bin/python -c "import setuptools, tokenize;__file__='/tmp/pip_build_ubuntu/pandas/setup.py';exec(compile(getattr(tokenize, 'open', open)(__file__).read().replace('\r\n', '\n'), __file__, 'exec'))" install --record /tmp/pip-DxAgzf-record/install-record.txt --single-version-externally-managed --compile failed with error code 1 in /tmp/pip_build_ubuntu/pandas
Traceback (most recent call last):
  File "/home/ubuntu/.pyenv/versions/2.7.8/bin/pip", line 11, in <module>
    sys.exit(main())
  File "/home/ubuntu/.pyenv/versions/2.7.8/lib/python2.7/site-packages/pip/__init__.py", line 185, in main
    return command.main(cmd_args)
  File "/home/ubuntu/.pyenv/versions/2.7.8/lib/python2.7/site-packages/pip/basecommand.py", line 161, in main
    text = '\n'.join(complete_log)
UnicodeDecodeError: 'ascii' codec can't decode byte 0xe2 in position 72: ordinal not in range(128)
☁☀☂ setup.sh: ☁☀☂ 39: can't install optional for Python 2.7.8.


bash circle/setup.sh returned exit code 1

Action failed: bash circle/setup.sh

theengineear · 2016-11-01T16:22:13Z

Hmm. I think the tests need some attention in general. I'll come back to this later in the day and try to fix up our test suite. That is an old pip error I've seen before (not super unicode-friendly). I thought it was related to http://stackoverflow.com/questions/25036897/pip-install-unicodedecodeerror, but i don't know why it would have been in a directory with a unicode-y name.

At any rate, I'll take a look later today.

theengineear · 2016-11-01T16:23:29Z

Changes look good! I'll try to get a patch in to fix the test suite and we can get this in soon!

choldgraf · 2016-11-02T17:03:24Z

hey all - shall I take a look at this now to see if it works?

Kully · 2016-11-02T17:26:10Z

hey all - shall I take a look at this now to see if it works?

Go ahead! I just rebuilt the test and it failed again.

choldgraf · 2016-11-04T16:35:49Z

ok, will try to spend some time on this over the weekend

On Wed, Nov 2, 2016 at 10:26 AM Adam Kulidjian [email protected]
wrote:

hey all - shall I take a look at this now to see if it works?

Go ahead! I just rebuilt the test and it failed again.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#599 (comment), or mute
the thread
https://github.com/notifications/unsubscribe-auth/ABwSHUOj27e_T7e2zmRVsP5cAayp8xCIks5q6Me4gaJpZM4KmLGE
.

choldgraf · 2016-11-10T18:58:51Z

I can confirm that iplot now works on this branch. Still slow as heck but it does plot :)

Kully · 2016-11-10T19:09:23Z

I can confirm that iplot now works on this branch. Still slow as heck but it does plot :)

Do you know which PR solved it? I was working on one but the tests were not passing for some reason...

choldgraf · 2016-11-10T19:13:05Z

I just pulled the code on this PR and tried it out

Kully · 2016-11-10T19:15:21Z

Oh wait, I'm ridiculous. This PR was the PR I was talking about.

Yeah, it works but the tests are still failing for some reason... 😕

Kully · 2016-11-18T18:52:57Z

@choldgraf @theengineear Working on this now.

Finally getting test failures to occur locally - - helpful!

…e - '/' becomes ':'

Kully · 2016-11-21T23:30:16Z

Getting there. Cleared up this error which #614 was getting:

Kully · 2016-11-22T17:00:11Z

@chriddyp Can I get a 💃 ? Finally we have success.

Who else should I ping in the future? Is one reliable person enough? 😄

chriddyp · 2016-11-22T17:02:47Z

Who else should I ping in the future?

Me or @theengineear !

Kully · 2016-11-22T17:03:58Z

Me or @theengineear !

K, sounds good. Let me know if you approve here.

chriddyp · 2016-11-22T17:08:02Z

plotly/plotly/plotly.py

@@ -869,7 +868,7 @@ def _fill_in_response_column_ids(cls, request_columns,
        for req_col in request_columns:
            for resp_col in response_columns:
                if resp_col['name'] == req_col.name:
-                    req_col.id = '{0}/{1}'.format(grid_id, resp_col['uid'])
+                    req_col.id = '{0}:{1}'.format(grid_id, resp_col['uid'])


So did making plots based off of grid columns just not work before? Can you show me an example where this didn't work before but now, with this fix, it works?

Yeah, it wasn't working until I made the switch. Here's one example that didn't work before the change that worked after:

c1 = Column([6, 6, 6, 5], 'pillow') c2 = Column(['a', 'b', 'c', 'd'], 'couch2') g = Grid([c1, c2]) py.grid_ops.upload(g, 'pillow_and_couch_grid222', auto_open=False) url = py.plot([Scatter(xsrc=g[0], ysrc=g[1])], auto_open=False, filename='222pillow222')

I'm curious about something though...when plotly makes the request to V2 for the upload of the grid, prior to that, the local grid has blank ids (eg. grid[0].id returns "" ) and after the request, your local grid has the proper filled in id eg. "AdamKulidjian:1782:u3fg1"

How is my local grid getting updated if only a url after py.upload() is run?

Got it. That's weird, because this definitely worked at some point. Here is the section and example in our docs:

If this is the case, then it seems like maybe the plotly backend API changed and stopped allowing IDs in this format.

How is my local grid getting updated if only a url after py.upload() is run?

py.grid_ops.upload(g) updates the columns in g. Those columns in g have the same reference as c1 and c2, so changing them in g changes c1 and c2` too. See http://nedbatchelder.com/text/names.html for some more info.

The ids themselves are assigned by the plotly server.

chriddyp · 2016-11-22T17:09:28Z

plotly/tools.py

@@ -6208,6 +6207,9 @@ def create_dendrogram(X, orientation="bottom", labels=None,
        if len(s) != 2:
            exceptions.PlotlyError("X should be 2-dimensional array.")

+        if distfun is None:
+            distfun = scs.distance.pdist
+


what is the motivation behind this change? It seems like by removing it from the call signature it is less informative to users about what distfun is.

Perhaps, but for some strange reason the tests were failing here when scs() functions were in the calls. By putting them in the body of the functions it cleared up the problem.

I thought it had something to do with the order of the

try: import scipy as scp _scipy_imported = True except ImportError: _scipy_imported = False

or such but it wasn't related

Got it. should scs by scp in this case? Also, what happens if a user runs this function but they don't have scipy installed? It's not a required dependency of the plotly package (and it shouldn't be) but folks might be surprised if running the function fails due to a missing module.

What do we do in other parts of the code that use scipy? Do we just let it fail or do we tell users that they need to install scipy?

That's a good question. I feel like the error messages may do that enough. I mean, scipy is pretty commonplace as well. For this commit at least I don't think we should add the error messages, as that may be a big project in all parts of plotly.py

Right. It's just that the error won't be "scipy is not defined" because it won't break on the import scipy step, it'll break on the scs.distance.pdist step which means that the error will probably be scs is not defined which isn't very intuitive or helpful.

Could you try creating a clean virtualenv without scipy and try calling this function and seeing what the error message is? Scipy may be common but Plotly may be the first package and experience that users have in Python and we want to make it as easy as possible for people to get going. We can fix the issue in a separate PR but it'd be nice to know how this will break for our users.

Okay sounds good.

Fantastic! It does have a you-need-scipy message in it:

Great, thanks for looking into that!

chriddyp · 2016-11-22T17:12:36Z

Just left a couple of questions that I'd like to understand better before we merge. Otherwise looks good!

Kully · 2016-11-22T19:13:02Z

If this is the case, then it seems like maybe the plotly backend API changed and stopped allowing IDs in this format.

That's what I suspected, but wasn't prepared to dig into streambed ;)

chriddyp · 2016-11-22T23:49:21Z

OK, this looks good to me. 💃

Kully added 2 commits November 1, 2016 10:39

rewrote the args that go into _plot_html

756cfff

PEP-8'ed _iplot_html and removed commented out code

64f4cf9

updated default schema

ee7b976

Kully mentioned this pull request Nov 3, 2016

fixing config bug in #589 #601

Closed

updated default json schema

9c26e6c

Kully added 5 commits November 18, 2016 15:32

replaced / to : in (x/y)src

7e69cd9

some stuff

153f65b

added config - fixed offline

c6e7faf

something

32843af

updated the format plotly changes grid column ids after upload is mad…

63500ac

…e - '/' becomes ':'

Kully added 6 commits November 22, 2016 10:37

put scipy.stats in first scipy try in tools.py

9613bc9

schema + changed dendogram call

0798cf6

moved scs down a few lines

72c4030

keep doing things

b045ba9

added if None line for distfun

7c3aaf4

put import trys back to the way they were originally

ae9f339

chriddyp reviewed Nov 22, 2016

View reviewed changes

Kully merged commit 046ae17 into master Nov 22, 2016

Kully deleted the fix-iplot-offline branch November 22, 2016 23:50

cldougl mentioned this pull request Nov 29, 2016

offline plotting error when using domain other than plot.ly #621

Closed

pyup-bot mentioned this pull request May 25, 2017

Update plotly to 2.0.8 boazmohar/spineimageing#4

Merged

This was referenced Jun 30, 2017

Initial Update mozilla/emr-bootstrap-spark#84

Closed

Update plotly to 2.0.11 mozilla/emr-bootstrap-spark#92

Closed

Update plotly to 2.0.12 mozilla/emr-bootstrap-spark#95

Closed

This was referenced Aug 5, 2017

Update plotly to 2.0.13 mozilla/emr-bootstrap-spark#102

Closed

Update plotly to 2.0.14 mozilla/emr-bootstrap-spark#108

Closed

pyup-bot mentioned this pull request Aug 23, 2017

Update plotly to 2.0.15 mozilla/emr-bootstrap-spark#116

Closed

This was referenced Oct 6, 2017

Update plotly to 2.0.16 mozilla/emr-bootstrap-spark#121

Closed

Update plotly to 2.1.0 mozilla/emr-bootstrap-spark#122

Closed

This was referenced Oct 26, 2017

Update plotly to 2.2.0 mozilla/emr-bootstrap-spark#123

Closed

Update plotly to 2.2.1 mozilla/emr-bootstrap-spark#125

Closed

pyup-bot mentioned this pull request Nov 24, 2017

Update plotly to 2.2.2 mozilla/emr-bootstrap-spark#129

Closed

pyup-bot mentioned this pull request Dec 5, 2017

Update plotly to 2.2.3 mozilla/emr-bootstrap-spark#135

Closed

pyup-bot mentioned this pull request Jan 26, 2018

Update plotly to 2.3.0 mozilla/emr-bootstrap-spark#192

Closed

This was referenced Feb 16, 2018

Update plotly to 2.4.0 mozilla/emr-bootstrap-spark#211

Closed

Update plotly to 2.4.1 mozilla/emr-bootstrap-spark#218

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix iplot offline #599

Fix iplot offline #599

Kully commented Nov 1, 2016

Kully commented Nov 1, 2016

Kully commented Nov 1, 2016

theengineear commented Nov 1, 2016

theengineear commented Nov 1, 2016

choldgraf commented Nov 2, 2016

Kully commented Nov 2, 2016

choldgraf commented Nov 4, 2016

choldgraf commented Nov 10, 2016

Kully commented Nov 10, 2016

choldgraf commented Nov 10, 2016

Kully commented Nov 10, 2016

Kully commented Nov 18, 2016

Kully commented Nov 21, 2016

Kully commented Nov 22, 2016

chriddyp commented Nov 22, 2016

Kully commented Nov 22, 2016

chriddyp Nov 22, 2016

Kully Nov 22, 2016

chriddyp Nov 22, 2016 •

edited

Loading

chriddyp Nov 22, 2016 •

edited

Loading

chriddyp Nov 22, 2016

Kully Nov 22, 2016 •

edited

Loading

chriddyp Nov 22, 2016

Kully Nov 22, 2016

chriddyp Nov 22, 2016 •

edited

Loading

Kully Nov 22, 2016

Kully Nov 22, 2016

chriddyp Nov 22, 2016

chriddyp commented Nov 22, 2016

Kully commented Nov 22, 2016

chriddyp commented Nov 22, 2016

Fix iplot offline #599

Fix iplot offline #599

Conversation

Kully commented Nov 1, 2016

Kully commented Nov 1, 2016

Kully commented Nov 1, 2016

theengineear commented Nov 1, 2016

theengineear commented Nov 1, 2016

choldgraf commented Nov 2, 2016

Kully commented Nov 2, 2016

choldgraf commented Nov 4, 2016

ok, will try to spend some time on this over the weekend

choldgraf commented Nov 10, 2016

Kully commented Nov 10, 2016

choldgraf commented Nov 10, 2016

Kully commented Nov 10, 2016

Kully commented Nov 18, 2016

Kully commented Nov 21, 2016

Kully commented Nov 22, 2016

chriddyp commented Nov 22, 2016

Kully commented Nov 22, 2016

Choose a reason for hiding this comment

Choose a reason for hiding this comment

chriddyp Nov 22, 2016 • edited Loading

Choose a reason for hiding this comment

chriddyp Nov 22, 2016 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Kully Nov 22, 2016 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

chriddyp Nov 22, 2016 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

chriddyp commented Nov 22, 2016

Kully commented Nov 22, 2016

chriddyp commented Nov 22, 2016

chriddyp Nov 22, 2016 •

edited

Loading

chriddyp Nov 22, 2016 •

edited

Loading

Kully Nov 22, 2016 •

edited

Loading

chriddyp Nov 22, 2016 •

edited

Loading