output of SlidingWindowFeature and SlidingWindow seems incorrect

[amitli1]

I'm trying to understand how SlidingWindowFeature and SlidingWindow works.
I read the documentation but still confused with the results.

• Let's say I have data with 200 frames (i.e 200 ms) and each frame (i.e each 1-ms) have 4 features.
• Now I want to create SlidingWindow with duration of 100 ms and step of 50 ms
• I guess I should get the output of 3 windows:
  [0-100], [50-150], [100-200]

But when I run the following code, I'm getting 200 features:

from pyannote.core import SlidingWindowFeature, SlidingWindow
frame    = SlidingWindow(start=0.0, step=0.05, duration=0.100)
data     = np.random.randn(200, 4)
features = SlidingWindowFeature(data, frame)
features.getDimension(), features.getNumber()

output:
4, 200

• What am I missing ?
• Why I dont get 3 windows ?