I need to calculate moving average of a sensor data that is coming on on the serial port with Python. All the samples I can find about numpy use data from a file or hard coded data in an array before the program starts.
In my case I do not have any data when the program starts. The data comes in real time in every second. I want to smooth the data as it arrives on the serial port.
I have this working on the Arduino but also need it in Python. Can somebody please point me to a real time (single value over time) sample not a batch sample.
Here's how you would add one reading at a time to a running collection of readings and return the average. I prepopulated the readings list to show it in action, but in your program, you'd just start off with an empty list: readings = []
I made the assumption that you want to include the last x readings in your average rather than including all of the readings. That is what the max_samples parameter is for.
without numpy:
readings = [1, 2, 3, 4, 5, 6, 7, 8, 9]
reading = 10
max_samples = 10
def mean(nums):
return float(sum(nums)) / max(len(nums), 1)
readings.append(reading)
avg = mean(readings)
print 'current average =', avg
print 'readings used for average:', readings
if len(readings) == max_samples:
readings.pop(0)
print 'readings saved for next time:', readings
result:
current average = 5.5
readings used for average: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
readings saved for next time: [2, 3, 4, 5, 6, 7, 8, 9, 10]
with numpy:
import numpy as np
readings = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9])
reading = 10
max_samples = 10
readings = np.append(readings, reading)
avg = np.mean(readings)
print 'current average =', avg
print 'readings used for average:', readings
if len(readings) == max_samples:
readings = np.delete(readings, 0)
print 'readings saved for next time:', readings
result:
current average = 5.5
readings used for average: [ 1 2 3 4 5 6 7 8 9 10]
readings saved for next time: [ 2 3 4 5 6 7 8 9 10]