How to keep the first window constant in sliding window?

Question

I am using the following code to apply a sliding window on time-series data. I want to set up my first window as fixed and then apply the sliding window as shown below in the desired output.

df = pd.DataFrame({'B': [0, 1, 2, 3, 4, 5, 6,7,8,9,10]})
def sliding_window(data, size):
 return [ data[x:x+size] for x in range( len(data) - size + 1 ) ]
sliding_window(df, 7)

output

[ B
 0 0
 1 1
 2 2
 3 3
 4 4
 5 5
 6 6,
 B
 1 1
 2 2
 3 3
 4 4
 5 5
 6 6
 7 7,
 B
 2 2
 3 3
 4 4
 5 5
 6 6
 7 7
 8 8,
 B
 3 3
 4 4
 5 5
 6 6
 7 7
 8 8
 9 9,
 B
 4 4
 5 5
 6 6
 7 7
 8 8
 9 9
 10 10]

Desired output Example:

I am using the fixed window of size 5 here. and it always should be the first window, and the sliding window is same as before except it slides from first window. Like the left figure in the images.. enter image description here

[ B
 0 0
 1 1
 2 2
 3 3
 4 4,
 B
 0 0
 1 1
 2 2
 3 3
 4 4
 5 5,
 B
 0 0
 1 1
 2 2
 3 3
 4 4
 5 5
 6 6,
 B
 0 0
 1 1
 2 2
 3 3
 4 4
 5 5
 6 6
 7 7,
 B
 0 0
 1 1
 2 2
 3 3
 4 4
 5 5
 6 6
 7 7
 8 8,
 B
 0 0
 1 1
 2 2
 3 3
 4 4
 5 5
 6 6
 7 7
 8 8
 9 9,
 B
 0 0
 1 1
 2 2
 3 3
 4 4
 5 5
 6 6
 7 7
 8 8
 9 9
 10 10]

• ForceBru: Your desired output is not a sliding window though
• ferrelwill: Why do you think this is not a sliding window? Is the right part in the figure is a sliding window?
• ForceBru: the right part definitely is. Looks like the left part is called "expanding window".
• ForceBru: Your sliding_window function already implements this
• ForceBru: Yes, absolutely correct

Answer 1

Try this:

def rolling_window_maybe(data, initial_size: int):
 return [ data[:initial_size + x] for x in range( len(data) - initial_size ) ]

For example:

data = [1,2,3,4]
size = 2
data[:size + 0] == [1,2]
data[:size + 1] == [1,2,3]
data[:size + 2] == [1,2,3,4]