BUG: DataFrame.merge(suffixes=) does not respect None #24819

charlesdong1991 · 2019-01-17T20:28:06Z

closes BUG: DataFrame.merge(suffixes=) does not respect None #24782
tests added / passed
passes git diff upstream/master -u -- "*.py" | flake8 --diff
whatsnew entry

pep8speaks · 2019-01-17T20:28:19Z

Hello @charlesdong1991! Thanks for updating the PR.

Cheers ! There are no PEP8 issues in this Pull Request. 🍻

Comment last updated on February 04, 2019 at 17:16 Hours UTC

codecov · 2019-01-17T21:04:04Z

Codecov Report

Merging #24819 into master will decrease coverage by 49.47%.
The diff coverage is 0%.

@@             Coverage Diff             @@
##           master   #24819       +/-   ##
===========================================
- Coverage   92.38%    42.9%   -49.48%     
===========================================
  Files         166      166               
  Lines       52379    52385        +6     
===========================================
- Hits        48392    22478    -25914     
- Misses       3987    29907    +25920

Flag	Coverage Δ
#multiple	`?`
#single	`42.9% <0%> (-0.01%)`	⬇️

Impacted Files	Coverage Δ
pandas/core/reshape/merge.py	`9.43% <0%> (-84.84%)`	⬇️
pandas/io/formats/latex.py	`0% <0%> (-100%)`	⬇️
pandas/core/categorical.py	`0% <0%> (-100%)`	⬇️
pandas/io/sas/sas_constants.py	`0% <0%> (-100%)`	⬇️
pandas/tseries/plotting.py	`0% <0%> (-100%)`	⬇️
pandas/tseries/converter.py	`0% <0%> (-100%)`	⬇️
pandas/io/formats/html.py	`0% <0%> (-99.35%)`	⬇️
pandas/core/groupby/categorical.py	`0% <0%> (-95.46%)`	⬇️
pandas/io/sas/sas7bdat.py	`0% <0%> (-91.17%)`	⬇️
pandas/io/sas/sas_xport.py	`0% <0%> (-90.15%)`	⬇️
... and 124 more

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 512830b...cbb36b2. Read the comment docs.

codecov · 2019-01-17T21:04:05Z

Codecov Report

Merging #24819 into master will decrease coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master   #24819      +/-   ##
==========================================
- Coverage   92.37%   92.37%   -0.01%     
==========================================
  Files         166      166              
  Lines       52420    52406      -14     
==========================================
- Hits        48423    48409      -14     
  Misses       3997     3997

Flag	Coverage Δ
#multiple	`90.79% <100%> (-0.01%)`	⬇️
#single	`42.87% <0%> (-0.02%)`	⬇️

Impacted Files	Coverage Δ
pandas/core/reshape/merge.py	`94.48% <ø> (ø)`	⬆️
pandas/core/internals/managers.py	`96.06% <100%> (-0.01%)`	⬇️
pandas/core/common.py	`98.37% <0%> (-0.04%)`	⬇️
pandas/core/frame.py	`96.81% <0%> (-0.02%)`	⬇️
pandas/io/formats/html.py	`99.34% <0%> (-0.01%)`	⬇️
pandas/core/accessor.py	`98.79% <0%> (ø)`	⬆️
pandas/core/ops.py	`94.28% <0%> (ø)`	⬆️
pandas/core/internals/blocks.py	`94.17% <0%> (ø)`	⬆️
pandas/core/sparse/scipy_sparse.py	`100% <0%> (ø)`	⬆️
pandas/core/generic.py	`96.63% <0%> (ø)`	⬆️
... and 6 more

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update bb43726...71729b2. Read the comment docs.

jschendel

Thanks! Can you add a whatsnew note?

pandas/core/reshape/merge.py

pandas/tests/reshape/merge/test_merge.py

jreback

There is a bug here, where None in a tuple is coerced to a string, however, I am a bit iffy about automagically coercing None to empty string. Yes its convient, but if you have non-string labels then it is unexpected.

jreback · 2019-01-18T12:33:39Z

pandas/tests/reshape/merge/test_merge.py

+    b = pd.DataFrame({col2: [4, 5, 6]})
+
+    df = a.merge(b, left_index=True, right_index=True, suffixes=suffixes)
+    assert df.columns.tolist() == expected_cols


fully construct the result frame here

thanks, will change

jreback · 2019-01-18T12:40:41Z

pandas/core/reshape/merge.py

        self.right_on = com.maybe_make_list(right_on)

        self.copy = copy
+


does this not break if suffixes=None is passed?

it will indeed break, but do we expect to get None for suffixes? the default is ("_x", "_y"). I could understand that people may want to use suffixes like: (None, "_y") if they want to keep the name of one of original columns... but use case for suffixes=None seems rare...

jreback · 2019-01-18T12:41:29Z

doc/source/whatsnew/v0.24.0.rst

 - :meth:`Series.unstack` and :meth:`DataFrame.unstack` no longer convert extension arrays to object-dtype ndarrays. Each column in the output ``DataFrame`` will now have the same dtype as the input (:issue:`23077`).
 - Bug when grouping :meth:`Dataframe.groupby()` and aggregating on ``ExtensionArray`` it was not returning the actual ``ExtensionArray`` dtype (:issue:`23227`).
 - Bug in :func:`pandas.merge` when merging on an extension array-backed column (:issue:`23020`).
+- Bug in :func:`pandas.merge` when setting None in suffixes (:issue: `24782`).


be more explicit on what is changing here

thanks, I will change later once collecting more feedbacks from other committers. @jreback

jreback · 2019-01-18T12:42:42Z

@TomAugspurger @jorisvandenbossche

TomAugspurger

The docstring needs to be updated. Currently it's

suffixes : tuple of (str, str), default ('_x', '_y')
    Suffix to apply to overlapping column names in the left and right
    side, respectively. To raise an exception on overlapping columns use
    (False, False).

That'll need to be something like tuple of (str or None, str or None).

Even that type isn't quite right... Since (False, False) is apparently valid.

TomAugspurger · 2019-01-18T16:58:28Z

pandas/tests/reshape/merge/test_merge.py

+
+@pytest.mark.parametrize("col1, col2, suffixes, expected_cols", [
+    (0, 0, ("", "_dup"), ["0", "0_dup"]),
+    (0, 0, (None, "_dup"), ["0", "0_dup"]),


I don't think this get's the expected output of @simonjayhawkins in #24782, does it?

IIUC, @simonjayhawkins expected for (None, '_dup') to not cast the original input to a string. So the expected columns would be [0, "0_dup"]. Is that correct?

Even that type isn't quite right... Since (False, False) is apparently valid.

Yeah, it looks like pretty much anything is actually valid right now, e.g.

In [8]: df.merge(df, left_index=True, right_index=True, suffixes=(np.sum, pd.DataFrame)) Out[8]: 0<function sum at 0x000002550602C840> 0<class 'pandas.core.frame.DataFrame'> 0 1 1 In [9]: df.merge(df, left_index=True, right_index=True, suffixes=(list('abc'), pd.Series(list('xyz')))) Out[9]: 0['a', 'b', 'c'] 00 x\n1 y\n2 z\ndtype: object 0 1 1 1 1

Looks to be caused by items_overlap_with_suffix just hitting whatever is passed with a format:

pandas/pandas/core/internals/managers.py

Lines 1959 to 1985 in 08f92c4

def items_overlap_with_suffix(left, lsuffix, right, rsuffix):

"""

If two indices overlap, add suffixes to overlapping entries.

If corresponding suffix is empty, the entry is simply converted to string.

"""

to_rename = left.intersection(right)

if len(to_rename) == 0:

return left, right

else:

if not lsuffix and not rsuffix:

raise ValueError('columns overlap but no suffix specified: '

'{rename}'.format(rename=to_rename))

def lrenamer(x):

if x in to_rename:

return '{x}{lsuffix}'.format(x=x, lsuffix=lsuffix)

return x

def rrenamer(x):

if x in to_rename:

return '{x}{rsuffix}'.format(x=x, rsuffix=rsuffix)

return x

return (_transform_index(left, lrenamer),

_transform_index(right, rrenamer))

IIUC, @simonjayhawkins expected for (None, '_dup') to not cast the original input to a string.

Yeah, I overlooked this earlier but agree; it makes sense to me that None should not cast to string and instead retain the original type of the column label in question.

thanks! @TomAugspurger @jschendel i overlooked it... i changed the code a bit, and added several more test cases
And @jschendel I also feel like we should add some restrictions to the suffix type, apparently, it looks wired if any type is given as suffixes... is it worth it if i open another PR to address it?

simonjayhawkins · 2019-01-20T08:35:40Z

pandas/tests/reshape/merge/test_merge.py

+    expected = pd.DataFrame([[1, 4], [2, 5], [3, 6]],
+                            columns=expected_cols)
+
+    result = a.merge(b, left_index=True, right_index=True, suffixes=suffixes)


many tests in test_merge.py repeat using both the left.merge(right, **kwargs) and the pd.merge(left, right, **kwargs) syntax.

i'm not sure how important that is, so unless one of the maintainers comments otherwise this lgtm

thanks for your feedback, @simonjayhawkins ! if one of maintainers brings this up and wants it to be standardised in this PR, i will then fix it! otherwise, as you said, I will open another PR to tackle it!

simonjayhawkins · 2019-01-20T08:48:44Z

does this not break if suffixes=None is passed?

i guess that since duplicate labels are allowed in an index then suffixes=None should be a valid option. future PR though.

TomAugspurger · 2019-01-21T21:22:29Z

pandas/core/reshape/merge.py

    suffixes : 2-length sequence (tuple, list, ...)
        Suffix to apply to overlapping column names in the left and right
-        side, respectively
+        side, respectively. Except for tuple of (str, str), it also allows


I don't understand this. Is suffixes=[None, None] valid?

Would the following be correct?

Length-2 sequence. Each element should be a str or None. * str : the string is appended to the overlapping column label. * None : the column label is left as-is.

Ah, I see now that it's invalid, because we won't return a frame with duplicate labels. But presumably, a [None, 'a'] is valid (could you add a test for that?). So perhaps in the docstring note that at least one of the values must not be None.

thanks Tom for your comment! @TomAugspurger
you are right, a (None, None) will raise an error which was in the code already. And a (None, 'a') will be valid, and None refers to the column label is left as it is... I think i already had this (None, 'a') in the pytest file. Shall I add another one?

I didn't see a test where suffixes was a list like [None, 'a'], just tuples.

ahh, sorry, i misunderstood... my bad...
Thx for your review! tests added! And also docstring is updated! @TomAugspurger

Ah, I see now that it's invalid, because we won't return a frame with duplicate labels

So perhaps in the docstring note that at least one of the values must not be None.

@TomAugspurger Not following you there... we do return frames with duplicate labels if 1) we don't pass "suffixes", or 2) we pass "suffixes" such that they result duplicated labels. So I don't see anything offensive in having suffixes=(None, None) result in duplicated labels.

EDIT: had missed @simonjayhawkins 's comment below, stating basically the same thing.

TomAugspurger · 2019-01-21T21:56:42Z

pandas/core/reshape/merge.py

        Suffix to apply to overlapping column names in the left and right
-        side, respectively. Except for tuple of (str, str), it also allows
-        tuple of (None, str) or (str, None)
+        side, respectively. And each element should be either a str or None,


Sorry, this is still a bit confusing, and I don't think listing each combination really helps. How about.

suffixes : Sequence A length-2 sequence where each element is a optionally a string indicating the suffix to add to overlapping column names in `left` and `right` respectively. Pass a value of `None` instead of a string to indicate that the column names from `left` or `right` should be left as-is, with no suffix. At least one of the values must not be None.

Thanks, Tom! and sorry for my bad docstring! and yours does look much clearer! I just removed a redundant 'a' ^^ @TomAugspurger

simonjayhawkins · 2019-01-21T22:13:04Z

pandas/core/reshape/merge.py

-    suffixes : 2-length sequence (tuple, list, ...)
-        Suffix to apply to overlapping column names in the left and right
-        side, respectively
+    suffixes : Sequence


the docstring will also need the default values included

oh, yeah, added! thx for the carefulness!! @simonjayhawkins

simonjayhawkins · 2019-01-21T22:31:28Z

pandas/tests/reshape/merge/test_merge.py

+    ("a", "a", ("_x", None), ["a_x", "a"]),
+    ("a", "b", ("_x", None), ["a", "b"]),
+    ("a", "a", [None, "_x"], ["a", "a_x"]),
+    (0, 0, ["_a", None], ["0_a", 0])


for added assurance can you also add a couple of regression tests here

("a", "a", None, ["a_x", "a_y"]), (0, 0, None, ["0_x", "0_y"])

should be sufficient.

you mean ("a", "a", ("_x", "_y"), ["a_x", "a_y"]) probably? suffix cannot accept None for now. @simonjayhawkins

thanks! just added a small new test function to cover it!

does this not break if suffixes=None is passed?

i guess that since duplicate labels are allowed in an index then suffixes=None should be a valid option. future PR though.

suffixes=None currently raises TypeError: 'NoneType' object is not iterable.

Is suffixes=[None, None] valid?

suffixes=[None, None] currently raises ValueError: columns overlap but no suffix specified:

so i guess suffixes=None should be equivalent to suffixes=[None, None]. future PR thoiugh.

pandas/core/internals/managers.py

simonjayhawkins · 2019-01-22T00:35:27Z

pandas/tests/reshape/merge/test_merge.py

+
+@pytest.mark.parametrize("suffixes", [
+    (None, None),
+    ('', None),


with the changes you've made the left label would be cast to a string and the right label unchanged. so technically this shouldn't fail.

you are right, they are indeed different types, however, before applying the renamer function, there is a suffix check, which means None and '' cannot be assigned at the same time, so ('', None) will raise an error.

if not lsuffix and not rsuffix: raise ValueError('columns overlap but no suffix specified: ' '{rename}'.format(rename=to_rename))

Do you want it to be changed? @simonjayhawkins

the docstring reads At least one of the values must not be None. and ('',None) satisfies that.

Do you want it to be changed?

what would be the extent of the changes to make this work?

added already!
allow (None, '') for number column, and if column name is string, (None, '') or (None, None) will raise error

simonjayhawkins · 2019-01-22T14:30:59Z

pandas/core/reshape/merge.py

        `right` should be left as-is, with no suffix. At least one of the
-        values must not be None.
+        values must not be None. A combination of `''` and `None` is will
+        raise error for columns which type is string


my opinion is that adding A combination of `''` and `None` is will raise error for columns which type is string to the docstring is probably unnecessary.

it makes me wonder whether the (None,None) combination should be addressed within this PR after all. if the 'columns overlap but no suffix specified' check is removed, what breaks?

what do you mean (None, None) combination, i am a bit confused? we don't expect column names to be overlapped, do we?
I think now the suffixes takes in length-2 sequence correctly given different types of suffix, the only thing now is suffixes=None will raise an error, since now it requires a length-2 sequence, the code should slightly be changed in _MergeOperation class. But looks like it's out of scope of this PR although it's easy to fix...
I would advise to create another PR to address this issue, and i can work on this new PR to tackle it. and thanks for your review, look forward to hearing your opinion! @simonjayhawkins

what do you mean (None, None) combination

i mean the ability to return a DataFrame with duplicate column labels if so desired. i'm not sure why this isn't allowed. maybe it breaks things or maybe that check is not necessary.

if it was allowed would it make for a simpler api description in the docstring?

we could leave it for maintainer's discussion, if they want it, I will open a follow-up PR to address it.
Thanks for your feedback, and I just shortened the docstring! @simonjayhawkins

charlesdong1991 · 2019-01-27T11:35:08Z

@jreback @TomAugspurger @jschendel @simonjayhawkins I just moved whatsnew from 0.24.0 to 0.24.1 because the new version of Pandas had been released. Please take a look at this PR to see what else I should change. Thanks!

jreback · 2019-02-01T21:00:15Z

@charlesdong1991 this would be for 0.25.0 anyhow, pls merge master

charlesdong1991 · 2019-02-03T10:54:14Z

After discussing with @jorisvandenbossche , and as he suggested, this PR will only solve the case when at most one element is None, like (None, str) or (str, None) will remain column as-is when suffix is None. And for cases like (None, None) or None or False, (False, False), I will open a follow-up issue and PR to discuss and fix.

jorisvandenbossche · 2019-02-03T10:57:45Z

To be clear, the reason to leave out the (None, None) for now is that it already does something else: it does the same as (False, False) (raising an error if there will be duplicates). This is not documented or tested, so we could maybe just change it (or otherwise first deprecate it), but let's leave that for the follow-up issue for discussion.

charlesdong1991 · 2019-02-03T12:44:51Z

I just did another change based on what we discussed before i left the sprint @jorisvandenbossche , and @toobaz please also take a look to see if we could merge and close this, and then, I will open another issue and PR addressing the potential issue and change for suffixes.

jorisvandenbossche

Looks good! Some small comments.
Can you also update the docstring of pd.merge ?

jorisvandenbossche · 2019-02-03T15:32:43Z

pandas/tests/reshape/merge/test_merge.py

+    a = pd.DataFrame({col1: [1, 2, 3]})
+    b = pd.DataFrame({col2: [3, 4, 5]})
+
+    msg = "columns overlap but no suffix specified"


Can you add here a comment like # TODO reconsider current behaviour of raising, see #.... (with a link to the issue).
Just in case later on when we want to change this, we won't think like "oh no, we cannot change it, because it is tested behaviour")

thanks, added!

jorisvandenbossche · 2019-02-03T15:33:17Z

pandas/core/reshape/merge.py

-    suffixes : 2-length sequence (tuple, list, ...)
-        Suffix to apply to overlapping column names in the left and right
-        side, respectively
+    suffixes : Sequence or None, default is ("_x", "_y")


The or None can be removed I think (we removed that change from this PR)

jreback

lgtm. minor comments, ping on green.

jreback · 2019-02-04T13:23:45Z

doc/source/whatsnew/v0.25.0.rst

 Reshaping
 ^^^^^^^^^

+- Bug in :func:`pandas.merge` doesn't work correctly if None is in suffixes (:issue: `24782`).


can you be a bit more clear on what the previous sympton was, instead of 'doesn't work correctly'

double backticks on None

no space after the colon (:issue:`24782`)

jreback · 2019-02-04T13:24:31Z

pandas/core/internals/managers.py

-                return '{x}{lsuffix}'.format(x=x, lsuffix=lsuffix)
-            return x
+        def renamer(x, suffix):
+            """Rename the left and right indices.


can you make a proper doc-string (Parameters / Returns)

jreback · 2019-02-04T13:25:59Z

pandas/tests/reshape/merge/test_merge.py

+    (0, 0, [None, None]),
+    (0, 0, (None, ""))
+])
+def test_merge_error(col1, col2, suffixes):


test_merge_suffix_error

jreback · 2019-02-04T13:26:24Z

pandas/core/reshape/merge.py

+        `left` and `right` respectively. Pass a value of `None` instead
+        of a string to indicate that the column name from `left` or
+        `right` should be left as-is, with no suffix. At least one of the
+        values must not be None.


can you add a versionchanged 0.25.0 here

jreback · 2019-02-04T13:27:07Z

suffixes=None is still invalid right (e.g. passing an actual None value)? this has a test?

charlesdong1991 · 2019-02-04T15:33:33Z

thanks @jreback for your review... all changed based on review, pls let me know if you have other questions.

jreback · 2019-02-06T03:51:09Z

thanks @charlesdong1991

jorisvandenbossche · 2019-02-06T07:23:17Z

@charlesdong1991 do you want open a new issue for changing the (None, None) case and allowing None and False ?

charlesdong1991 · 2019-02-06T09:07:21Z

@jorisvandenbossche yes! I will open a new issue addressing None related case later today or tomorrow!!

simonjayhawkins · 2019-02-09T00:11:48Z

@charlesdong1991 : the Panel.join tests have now been removed, xref #25191. so there should no longer be issues with the code shared between Panel.join and DataFrame.merge with regard to the suffixes=(False, False) behaviour that was causing us problems at europandas.

charlesdong1991 · 2019-02-09T16:05:24Z

@simonjayhawkins thanks a lot!! I also noticed it when merging my branch to master!!

@jorisvandenbossche sorry that recently i was quite busy so didn't have time to work on new PR, i just created a new PR to address None/False behaviour.

jschendel reviewed Jan 17, 2019

View reviewed changes

pandas/core/reshape/merge.py Outdated Show resolved Hide resolved

pandas/tests/reshape/merge/test_merge.py Show resolved Hide resolved

pandas/tests/reshape/merge/test_merge.py Outdated Show resolved Hide resolved

jschendel added Bug Reshaping Concat, Merge/Join, Stack/Unstack, Explode labels Jan 17, 2019

jreback requested changes Jan 18, 2019

View reviewed changes

TomAugspurger reviewed Jan 18, 2019

View reviewed changes

simonjayhawkins reviewed Jan 20, 2019

View reviewed changes

TomAugspurger reviewed Jan 21, 2019

View reviewed changes

simonjayhawkins reviewed Jan 21, 2019

View reviewed changes

jschendel reviewed Jan 21, 2019

View reviewed changes

pandas/core/internals/managers.py Outdated Show resolved Hide resolved

jschendel reviewed Jan 21, 2019

View reviewed changes

pandas/core/internals/managers.py Outdated Show resolved Hide resolved

simonjayhawkins reviewed Jan 22, 2019

View reviewed changes

charlesdong1991 force-pushed the bug_suffix_none branch from d44cdf0 to ed91045 Compare January 27, 2019 11:26

one in all

82c52a4

charlesdong1991 force-pushed the bug_suffix_none branch from 583c722 to 82c52a4 Compare February 2, 2019 16:15

charlesdong1991 added 5 commits February 2, 2019 16:16

change whatsnew to 0.25

dd605e0

add back other comment

af7f9ad

double check test

4d5e1a9

changes based on discussion

3f65bf1

changes based on discussion with joris

ce7e4b8

slight change

90ca9cd

jorisvandenbossche reviewed Feb 3, 2019

View reviewed changes

slight changes

e995a04

jreback requested changes Feb 4, 2019

View reviewed changes

changes based on jeff review

9c3dfbd

charlesdong1991 added 2 commits February 4, 2019 17:34

fix test error

441e9a5

ci fail, try again

71729b2

jreback added this to the 0.25.0 milestone Feb 6, 2019

jreback approved these changes Feb 6, 2019

View reviewed changes

jreback merged commit 09633b8 into pandas-dev:master Feb 6, 2019

charlesdong1991 mentioned this pull request Feb 9, 2019

ENH: accept None behaviour for suffixes in DataFrame.merge #25242

Closed

simonjayhawkins mentioned this pull request May 16, 2022

Merge nonstring columns #46879

Closed

4 tasks

		self.right_on = com.maybe_make_list(right_on)

		self.copy = copy

	def items_overlap_with_suffix(left, lsuffix, right, rsuffix):
	"""
	If two indices overlap, add suffixes to overlapping entries.

	If corresponding suffix is empty, the entry is simply converted to string.

	"""
	to_rename = left.intersection(right)
	if len(to_rename) == 0:
	return left, right
	else:
	if not lsuffix and not rsuffix:
	raise ValueError('columns overlap but no suffix specified: '
	'{rename}'.format(rename=to_rename))

	def lrenamer(x):
	if x in to_rename:
	return '{x}{lsuffix}'.format(x=x, lsuffix=lsuffix)
	return x

	def rrenamer(x):
	if x in to_rename:
	return '{x}{rsuffix}'.format(x=x, rsuffix=rsuffix)
	return x

	return (_transform_index(left, lrenamer),
	_transform_index(right, rrenamer))

Uh oh!

BUG: DataFrame.merge(suffixes=) does not respect None #24819

BUG: DataFrame.merge(suffixes=) does not respect None #24819

Uh oh!

Conversation

charlesdong1991 commented Jan 17, 2019

Uh oh!

pep8speaks commented Jan 17, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Comment last updated on February 04, 2019 at 17:16 Hours UTC

Uh oh!

codecov bot commented Jan 17, 2019

Codecov Report

Uh oh!

codecov bot commented Jan 17, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

jschendel left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

jreback left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jreback commented Jan 18, 2019

Uh oh!

TomAugspurger left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

simonjayhawkins commented Jan 20, 2019

Uh oh!

Choose a reason for hiding this comment

Uh oh!

charlesdong1991 Jan 21, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

toobaz Feb 2, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

pep8speaks commented Jan 17, 2019 •

edited

Loading

codecov bot commented Jan 17, 2019 •

edited

Loading

charlesdong1991 Jan 21, 2019 •

edited

Loading

toobaz Feb 2, 2019 •

edited

Loading

charlesdong1991 Jan 22, 2019 •

edited

Loading

charlesdong1991 commented Jan 27, 2019 •

edited

Loading

charlesdong1991 commented Feb 3, 2019 •

edited

Loading