_repr_html_ returns None

[lbillingham]

In previous versions, the Spyder IDE displayed HTML for DataFrames in its iPython console. In 0.15.2 they are no longer displayed.

I think this may be to do with DataFrame._repr_html_()

If I run the following on 0.15.2:

import pandas as pd
import numpy as np

print('pandas version ', pd.__version__)
rng = pd.date_range(start='2014-01-01', periods=5, freq='D')
df = pd.DataFrame({'y':np.random.normal(size=len(rng))}, index=rng)
ret = df._repr_html_()
print(None == ret)

it prints 0.15.2, True

and viewing the DataFrame with df results in the text rendering:

On 0.14.1, running the above I get 0.14.1, False
and df results in this:

[shoyer]

Can you share pd.show_versions()?

[lbillingham]

Oh, is this maybe an IPython thing? There are different versions of that on the 2 boxen.

repr_html() == None

INSTALLED VERSIONS

commit: None
python: 2.7.7.final.0
python-bits: 64
OS: Windows
OS-release: 7
machine: AMD64
processor: Intel64 Family 6 Model 58 Stepping 9, GenuineIntel
byteorder: little
LC_ALL: None
LANG: en_GB

pandas: 0.15.2
nose: 1.3.4
Cython: 0.20.1
numpy: 1.9.1
scipy: 0.15.0
statsmodels: 0.5.0
IPython: 2.3.1
sphinx: 1.2.3
patsy: 0.2.1
dateutil: 1.5
pytz: 2014.9
bottleneck: None
tables: 3.1.1
numexpr: 2.3.1
matplotlib: 1.3.1
openpyxl: 1.8.5
xlrd: 0.9.3
xlwt: 0.7.5
xlsxwriter: 0.5.5
lxml: 3.3.5
bs4: 4.3.1
html5lib: None
httplib2: None
apiclient: None
rpy2: None
sqlalchemy: 0.9.4
pymysql: None
psycopg2: None

repr_html() != None

INSTALLED VERSIONS

commit: None
python: 2.7.8.final.0
python-bits: 32
OS: Windows
OS-release: 7
machine: AMD64
processor: Intel64 Family 6 Model 58 Stepping 9, GenuineIntel
byteorder: little
LC_ALL: None
LANG: None

pandas: 0.14.1
nose: 1.3.4
Cython: 0.21
numpy: 1.9.0
scipy: 0.14.0
statsmodels: 0.5.0
IPython: 2.2.0
sphinx: 1.2.3
patsy: 0.3.0
scikits.timeseries: None
dateutil: 1.5
pytz: 2014.7
bottleneck: None
tables: 3.1.1
numexpr: 2.3.1
matplotlib: 1.4.0
openpyxl: 1.8.5
xlrd: 0.9.3
xlwt: 0.7.5
xlsxwriter: 0.5.7
lxml: 3.4.0
bs4: 4.3.2
html5lib: None
httplib2: None
apiclient: None
rpy2: None
sqlalchemy: 0.9.7
pymysql: None
psycopg2: None

[shoyer]

Can you verify that you're using running the same version of Spyder in both environments?

I have two other things you can try that might give us a hint: what do you get from running pd.core.common.in_qtconsole() and pd.get_option("display.notebook_repr_html") ?

[jorisvandenbossche]

xref #7372

I also saw that Spyder now shows the text repr and no longer the html repr (and I think this is a good idea, see the issue I linked -> can lead to strange effects. And the ipython devs are even removing the default html displaying from the qtconsole in ipython 3.0, as I understand).
But for me, I thought this was due to upgrading my spyder from 2.3.0 to 2.3.2, so indeed, can you check your spyder versions?

[lbillingham]

They are running different versions of Spyder and in_qtconsole() has changed

###repr_html() == None
Spyder version 2.3.2

pd.get_option("display.notebook_repr_html")
True

pd.core.common.in_qtconsole()
True

###repr_html() != None
Spyder version 2.3.0

pd.get_option("display.notebook_repr_html")
True


pd.core.common.in_qtconsole()
False

I'd say this was closed.

I couldn't find anything in the release notes for Spyder about this. I've suggest adding something.
https://code.google.com/p/spyderlib/issues/detail?id=2126

Thank you for your help.

[jorisvandenbossche]

So to summarize: pandas has a in_qtconsole function that is used to not show the html repr in qtconsole as this is not done properly (see comment here: https://github.com/pydata/pandas/blob/v0.15.2/pandas/core/frame.py#L505).
Spyder changed some things in 2.3.2 to let the ipython console behave like a 'real' qtconsole for pandas in order to fix a bug (https://code.google.com/p/spyderlib/issues/detail?id=2015), but also with the result that now the text repr is used.

@ccordoba12 Maybe you can add this to your changelog?

[ccordoba12]

@jorisvandenbossche @lbillingham I didn't think this was important enough to merit a changelog note :-) I'll add one for our next release because I doubt people read old changelogs.

In a way this was not our fault. The problem was in_qtconsole was not doing their job correctly (by assuming that the only way to have a qtconsole is to run ipython qtconsole and detecting the application name). I just noticed it when a user reported our issue 2015

I hope you fix this behavior :-)