Running mean

Authors:Jose Guzman
Updated:20 February, 2019

The running mean (or running average) is simple way to smooth the data. Given a certain set of points, a running average will create a new set of data points which will be computed by adding a series of averages of different subsets of the full data set.

Given for example a sequence X of n points, we can create a new set of data points S of length n by taking the average of a subset of w points from the original data set for every point S_i within the set:

{\displaystyle S_i=\frac{1}{w} \sum^{w+i}_{j=i} X_j }

The running mean function

The following Python function calculates the running mean of the current channel. Both trace and channel can be selected as zero-based indices. The width of the running average (refereed to here as binwidth) can be selected. The resulting average will appear in a new Stimfit window.

# load main Stimfit module
import stf

# load NumPy for numerical analysis
import numpy as np

def rmean(binwidth, trace=-1,channel=-1):
    Calculates a running mean of a single trace


    binwidth    -- size of the bin in sampling points (pt).
    Obviously, it should be smaller than the length of the trace.

    trace:  -- ZERO-BASED index of the trace within the channel.
    Note that this is one less than what is shown in the drop-down box.
    The default value of -1 returns the currently displayed trace.

    channel  -- ZERO-BASED index of the channel. This is independent
    of whether a channel is active or not. The default value of -1
    returns the currently active channel.


    A smoothed traced in a new stf window.

    # loads the current trace of the channel in a 1D Numpy Array
    sweep = stf.get_trace(trace,channel)

    # creates a destination python list to append the data
    dsweep = np.empty((len(sweep)))

    # running mean algorithm
    for i in range(len(sweep)):

        if (len(sweep)-i) > binwidth:
            # append to list the running mean of `binwidth` values
            # np.mean(sweep) calculates the mean of list
            # sweep[p0:p10] takes the values in the vector between p0 and p10 (zero-based)
            dsweep[i] = np.mean( sweep[i:(binwidth+i)] )

            # use all remaining points for the average:
            dsweep[i] = np.mean( sweep[i:] )


Code commented

Stimfit commonly uses the value -1 to set the current trace/Channel. In this function the default argument values are -1 (see the function arguments trace=-1 and channel=-1).

>>> sweep = stf.get_trace(trace,channel)

stf.get_trace() imports the trace of the channel into a 1D-Numpy array that we called sweep. The default values provided by the function are -1. This means that by default, the current trace/channel will be imported.

We create a new stf window with the following

>>> stf.new_window(dsweep)

where dsweep is the 1D-NumPy array obtained after performing the running average.


To perform the running average of 10 sampling points of the current trace, type:

>>> spells.rmean(10)

A new window with the running mean will appear.