timeseries

5. Timeseries module

Learn how to use the timeseries object. The timeseries object is used for all timeseries related operations both filterbank and non-filterbank (general array) types.

5.1 Initialize the timeseries object

The Timeseries module can be used with both the filterbank file and a general numpy array.

5.1.1 Initialize using a filterbank file

• Create an filterbank object using the standard filterbank module.
• Read the filterbank file in memory (do not use streams)
• Initialize the timeseries object using the filterbank object.

import filterbank.filterbank as Filterbank
import timeseries.timeseries as Timeseries

# Read the filterbank file from a file. 
filterbank_obj = Filterbank('./pspm32.fil')

# Read the filterbank as a whole instead of as a stream. 
filterbank_obj =  filterbank_obj.read_filterbank()

# Initialize the timeseries object. 
ts = Timeseries().from_filterbank(filterbank_obj)

5.1.2 Initialize using a numpy array

import numpy as np
import timeseries.timeseries as Timeseries

input_array = np.array([1, 2, 3, 4, 5, 7, 9, 10, 11])

ts = Timeseries(input_array)

5.2 Retrieve the timeseries array from timeseries object

5.2.1 Retrieve timeseries object

# Assumed that your timeseries object has been initialized. 

timeseries_array = timeseries.get()

5.3 Downsample the timeseries

The first implemented feature for the timeseries object is the downsample/decimate function. This enables you to downsample your timeseries by q scale. This will make your input array smaller and basically 'cuts' the other parts off.

5.3.1 Downsample/Decimate

Called downsample in the current release because no anti-alliasing is used, might be renamed to decimate once more advanced operations (such as antialiasing) are used.

# Downsampled array shall be 3 times smaller than the current timeseries (as initialized) 
scale = 3
# Returns an array with the downsampled timeseries, can also be retreived lated user timseries.get()
downsampled_array = timeseries.downsample(scale)

Back to table of contents