Slightly idiosyncratic usage of groupby

[dhirschfeld]

When grouping by month in the notebook below you use a pd.Grouper object:

pbpython/notebooks/pandas-styling.ipynb

Lines 733 to 735 in d170f44

	"monthly_sales = df.groupby([pd.Grouper(key='date', freq='M')])['ext price'].agg(['sum']).reset_index()\n",
	"monthly_sales['pct_of_total'] = monthly_sales['sum'] / df['ext price'].sum()\n",
	"monthly_sales"

..whereas the same could be accomplished by simply using the resample method:

monthly_sales = df.resample('MS', on='date')['ext price'].agg(['sum'])

Whilst I think it's useful for power users to know they can construct Grouper objects themselves I think for less advanced users it's an internal implementation detail which they don't really need to know and which makes it seem more complicated than it needs to be

[chris1610]

Thanks for the feedback. It's a good point.

I'll update the notebook and article shortly.

Thank you for the feedback.

[dhirschfeld]

It's a great article! I'm planning on sharing it with my users, most of whom are just learning Python.