How to get the average of the last n rows into a new column in pandas?

Question

Python

Greetings. I'll write it in more detail, I don't know how to google it.

There is a table:

a = 
df = pd.DataFrame(a, columns=['low', 'high'])
df

No problem, it turns out to calculate the average between the columns

df = df.assign(hl2 = (df.low + df.high) / 2)
df

But how to calculate the average for each last 3 rows?
Tried like this:

df = df.assign(mean3 = df.tail(3).hl2.mean())
df

And I got the result everywhere in general for the last 3 lines


. You can, of course, sort through the loop, but with large data this is very inefficient. I think there must be some native method for this?

For example, the prometheus monitoring system uses a method in the rate request that does something similar.

Answer 1

Found the answer. rolling method

df = df.assign(mean3 = df.rolling(3, min_periods=1).hl2.mean())
df