I have a question to all you filter experts out there. I have some engine data sampling irregularly at about 50Hz. Recently I've been downsampling via the Pandas resample().mean() method. Basically I have a 50 Hz signal that I'd like to downsample to 1 Hz. What are the pitfalls of using something like this. It's not quite a rolling mean filter as it doesn't quite roll. As far as I understand, by taking the mean of the binned data, I'm introducing quite a bit of noise in the freq space. However, this method does handle NaN's in my data quite well (of which there are a lot). I just wanted to know peoples thoughts on using this.