pdutils - pandas utilities

Introduction

The intention of this package is simple - to make testing with pandas easy.

Here's an example of the intended API usage :

from pdutils import ts_compare, assert_

...

def test1():
    # initialise two pandas TimeSeries objects for comparison, then ...
    assert_(ts_compare(ts1, ts2))

The test should pass if both TimeSeries objects are equivalent.

If the comparison fails, by default, the assertion should provide a terse, yet meaningful message about the difference that cause the assertion failure.

A more details comparison can be extracted using the verbose option.

def test2():
    # initialise two pandas TimeSeries objects for comparison, then ...
    assert_(ts_compare(ts1, ts2, verbose=True))

This should provide an over and under view of the difference found in a particular column, where applicable.

For columns containing float types it is also possible to configure the tolerance of the comparison (see docstrings for further details).

API

Function exist to allow you to compare pandas TimeSeries, DataFrames and numpy arrays.

ts_compare(x, y)

Compares two pandas.TimeSeries object for equivalence.

df_compare(x, y)

Compares two pandas.DataFrame objects for equivalence.

ndarray_compare(x, y)

Compares two numpy arrays for equivalence.

assert_((expr, failure_msg))

Like the Python assert statement but in functional form.

assert_not((expr, failure_msg))

A negative assertion in functional form.