Using Unit Tests to Fix a Bug in Pandas

Using Unit Tests to Fix a Bug in Pandas


The issue that I decided to work on this week is one of the issues I mentioned two weeks ago when I was planning my work for the next couple of months. This time, I'm expanding on my knowledge of unit tests in Pandas from last week and using them to my advantage to fix an outstanding bug.

Overview of the Issue

The following issue has been posted on June 16, 2018. It's a pretty old issue and one person attempted to solve it. However, the moderators requested changes from the person working on the issue and he never committed anything that would satisfy their requests. After a while, one of the moderators stated that the pull request is now stale and closed it. 

On December 2, 2018 after ensuring that the issue was not yet resolved, one of the moderators updated the issue to have a "Contributions Welcome" milestone. Seeing that the moderators are still interested in a fix for this issue, I thought that it would be a good issue for me to work on.

At first, I was a little intimidated by "Difficulty Intermediate", but after looking further into the issue, I realized that there is already existing feedback that will help me get started on the issue without asking unnecessary questions.

The creator of the issue has brought attention to a bug that is caused when combining concat function and OrderedDict. OrderedDict is an ordered dictionary, which means that the dictionary should remember the order of the items added to this dictionary. However, when combining this dictionary with concat, the order gets lost. Here are the observations:

In [2]: from collections import OrderedDict

In [3]: pd.concat(OrderedDict([('First', pd.Series(range(3))),
   ...:                        ('Another', pd.Series(range(4)))]))
   ...:                        
Out[3]: 
Another  0    0
         1    1
         2    2
         3    3
First    0    0
         1    1
         2    2
dtype: int64

However, this should be the output:

First    0    0
         1    1
         2    2
Another  0    0
         1    1
         2    2
         3    3
dtype: int64


Writing the Unit Test

My first thought before attempting to fix this issue was to write a unit test that would test the above functionality. This would allow me to easily test the changes that I make to the code. With the experience that I acquired from my last pull request, where I wrote a unit test for Pandas, it was a lot easier to formulate one that would be good for this issue.

I followed the same principle of creating the buggy result and comparing it to the expected result, which I also had to recreate. You can observe the code of my test below.


When I first ran the following test, I got the following:


As you can see, the error that I get is expected. The left input is the result that I created based on the buggy behaviour and the right input is the expected behaviour. The following test matches the concern of the issue that I'm working on and would help me produce a fix.

If I manage to get the issue fixed, the unit test is also a valuable part of it because it will ensure that the bug doesn't occur again. Whenever something is added to the project, it is beneficial to also write a test for it, so other pull requests don't break the functionality that was just added.


Progress on the Bug

Now that the unit test is up and running, I started exploring ways to fix the actual bug. I looked at the pull request that was previously created to see what the previous person who attempted to work on this issue achieved. He achieved the expected behaviour and it allowed me to pass the test that I created, however the moderators requested changes on what he did, which were never addressed. Next week I will attempt to address these changes and get further feedback from the moderators once I create my pull request.

The following code was previously implemented to solve the bug:


The summary of change requests that followed:
  • In Python 3.7, regular dicts are also ordered and should be tested
  • However, implementation should separate 3.6 and 3.7. There are PY36 and PY37 constants that can be used for this check.
  • pandas.core.common._dict_keys_to_ordered_list should be refactored to help with this check.
Next week I will be using all this feedback and requests in order to come up with a solution.


Comments

Post a Comment

Popular posts from this blog

First Enhancement in Pandas

Image
First Enhancement in Pandas After spending the last few months fixing bugs in Pandas and writing tests, I decided that I wanted to try something a little bit different. I went on a hunt for enhancement issues to see if I can add another type of contribution to Pandas. A lot of enhancement requests that I found were hard and technical, so I stayed away from them for now. However, I found one that didn't require a whole new functionality to be added to the code, but asked to add an extra calculation in an existing function, and I proceeded to work on it. Issue https://github.com/pandas-dev/pandas/issues/21689 The issue deals with the pd.DataFrame.describe() function in Pandas. Currently, this function provides 8 summary statistics that are performed on the DataFrame that calls this function. These summary statistics include: count, mean, std, min, 25%, 50%, 75%, max. All of these summary statistics are useful in data analysis and provide a summary of the central t
Read more

Working with Incomplete MultiIndex keys in Pandas

Image
Incomplete MultiIndex keys in Dropna This week I produced a Pull Request for the bug in Pandas that I discussed last week. Although I think the solution that I provided can be further improved, it does fix the immediate issue of incomplete MultiIndex keys in the subset parameter of the "dropna()" function. I submitted a Pull Request to gather some feedback on my solution and improve it based on the moderator's discretion. Overview of the Issue https://github.com/pandas-dev/pandas/issues/17737 The creator of the issue brought to attention a bug that does not allow MultiIndex keys to be incomplete in the subset parameter of the "dropna()" function.  DataFrame.dropna(...) function removes missing values. If you pass in a "subset=..." parameter into it, it will remove rows based on the specified list of columns that are listed in the subset. I created some sample code, following the examples that the moderator gave and here a
Read more

Progress in Open Source

Image
Progress in Open Source In this blog I will talk about my journey and progress in open source development so far. I've been really enjoying working on real world projects that are used by many people every day. It is not only challenging as every Pull Request required me to perform research in order to understand the issue and solve it, but it is also rewarding because I acquire hands-on experience that is much needed and also contribute to the open source community that I am very fond of now. Open Source Timeline Summary I started exploring open source development during September of 2018. I had to learn a lot of useful tools that Git and Github offers, which now come in handy almost every day. I now know how to create meaningful issues that follow the specifications of the project, as well as create and structure my Pull Requests to the standards of the project that I'm working on. Furthermore learning how to properly manipulate version history using
Read more