Using Slice Objects in Python

Here's a cool little helpful piece of code I found for Python, it's called a slice object.

You know when you write my_list[:2] and you get the first two values? Well, the ":2" section is actually a slice object which you can input directly.

The notation for the object is slice(None, 2) or slice(None, 2, None), similar to when you subset a list using one or two colons.

Therefore, if we want to get the first two values of the list we can write:

my_list[slice(None, 2)]

Now, this doesn't actually save us anything from the usual syntax, but we can now save the slice object as its own variable:

first_two = slice(None,2)

my_list[first_two]

My most common use case for slice objects is with pandas dataframes.

In one case I was generating a report with many different subsets of a few tables. Sometimes I needed the last column, which was a 'total' column, and othertimes not.

Therefore, it was much easier to write:

def subset_my_df(df, include_total=True)
    '''This function takes a pandas dataframe and includes the 'total' column at the end, or not.'''
    if include_total:
        df_slice = slice(None,5)
    else:
        df_slice = slice(None,4)

    cols=df.columns[df_slice]

    subset_df = df[cols]

    return subset_df

Than to try to hardcode the columns without the total, especially if you have dataframes with different column names.

You need the cols=df.columns[df_slice] line to subset the columns. If you don't include that line you'll be subsetting by rows instead.