ashomah / ie_pandas Goto Github PK

I noticed that you are importing numpy inside the class definition:

https://github.com/lhandal/ie_pandas/blob/73253267fbaa5eb74dea3b12c73c1020b8eee88a/src/ie_pandas/__init__.py#L1-L3

However, it doesn't work as you expect:

juanlu@centauri /tmp $ cat importing.py 
class Foo:
    import numpy as np
    def foo(self):
        return np.arange(4)
    def correct_foo(self):
        return self.np.arange(4)
juanlu@centauri /tmp $ python3 -q
>>> from importing import Foo
>>> f = Foo()
>>> f.foo()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/tmp/importing.py", line 4, in foo
    return np.arange(4)
NameError: name 'np' is not defined
>>> f.correct_foo()
array([0, 1, 2, 3])

You still don't have tests for DataFrame (#5) and that's why you probably didn't notice so far.

Dear @Juanlu001,

we have a recurring problem during the testing. We want to test our DataFrame class for converting different input types with different value types in it against an expected_output print of the dataframe.

As an example, one of the pytests we made for this is the following

from ie_pandas import DataFrame
import pytest

def test_string_input_in_list():
    artists_age_list = ["James Blake", "Björk", "Christine & The Queens", "DJ Koze", "Solange"]

    df1 = DataFrame(artists_age_list)

    expected_output =                        0\n0             James Blake\n1                   Björk\n2  Christine & The Queens\n3                 DJ Koze\n4                 Solange
    
    actual_output = df1

    assert actual_output == expected_output

The problem situates itself on the expected_output line. We get the following error running pytest:

Also when we put quotation marks around the output, an Assertion error is risen.

How can we overcome this and make our test successful?

Thanks in advance!

It seems that you are all pushing branches to this repository, and that you are not using forks:

https://github.com/lhandal/ie_pandas/network/members

Even though some projects are carried out like this, I explicitly instructed you to work from forks:

"The library must be developed incrementally, using pull requests that should depart from a fork branch (i.e. not from master, not from a branch in the same repository)"

The reasons are mainly three:

• It's how you have to work on projects were you don't have write permissions (i.e. contributing to open source)
• Forces you to learn proper git workflow
• Removes the temptation to have several people pushing to the same branch, which is a recipe for disaster

Please keep this in mind from now on

Dear @Juanlu001 ,
I am a bit confused with why my commits are not counted in the contributors.

I think you previously commented on this, but I think I did not completely understand how to tackle this.

Thanks in advance for helping me out.

You are adding a lot of functionality that is not tested. I recommend that you

• Test what's already written
• Don't merge pull requests that add functionality without tests

Otherwise, there's a risk you leave the tests for the very end, and believe me - you will find last-minute problems that will make your life more difficult.

Hi @Juanlu001 ,
I'm reviewing our code using pycodestyle, and it returns many errors "E501 line too long".
How important are these ?
Our max line length is currently 88 characters.
Thank you !

Dear @Juanlu001,
when doing pytest --cov=ie_pandas tests/OK in our latest version with all working tests, seven tests are passed correctly and everything seems to be fine. However, all tests we wrote regarding datatypes seem to be ignored. Here you see an example of a test to see if a dict_of_lists with floats can be correctly converted to a df.

from ie_pandas import DataFrame
import pytest

def float_input_in_dict_of_lists():
    artists_dict_of_lists = {'age':[30.0, 53.0, 31.0, 47.0, 32.0], 'albums':[4.0, 10.0, 2.0, 5.0, 4.0]}
    dict_of_lists_df = DataFrame(artists_dict_of_lists, rowindex = ['R1', 'R2', 'R3', 'R4', 'R5'], colindex=['age', 'albums'])

    expected_output = "age  albums/nR1  30.0       4.0/nR2  53.0      10.0/nR3  31.0       2.0/nR4  47.0       5.0/nR5  32.0       4.0"

    actual_output = DataFrame(dict_of_lists_df)

    assert actual_output == expected_output

Can you tell us what should be rewritten to make this test passing as well?

Thanks in advance!

Even though you properly defined a .gitignore file, some .DS_Store files ended up in the repository. You should remove them.