I am trying to smoothen my time-series fluorescence intensity data, such as calcium imaging. I referred to the calculation method from the Box 1 of Jia et al., 2011, Nature protocol (doi:10.1038/nprot.2010.169), but there are some problems for me to follow.

  • They apply noise filtering by using exponentially weighted moving average, where the weight is different from those are able in the pandas.ewm method of python. Is there way to alter the weight in ewm method?
  • The author suggests parameters assuming that the data is taken 30 frames per sec (= 30 Hz), as extracted as follows."1. Typically, effective filtering for 30 Hz imaging is achieved by the following parameters: τ0 = 0.2 s, τ1 = 0.75 s and τ2 = 3 s." I could not find how to apply this when I have different frequency, do you know if there is equation or relationship for those parameters?
  • Other suggestions to analyze data would be helpful too!

    Sincerely, Park

    More Chanhyun Park's questions See All
    Similar questions and discussions