Visitar URL original
RecurrenceFilter: Apply an IIR filter to data—Wolfram Documentation

WOLFRAM

Wolfram Language & System Documentation Center

RecurrenceFilter[{α,β},x]

filters x using a linear recurrence equation with coefficients α and β.

RecurrenceFilter[tf,x]

uses a discrete-time filter defined by the TransferFunctionModel tf.

RecurrenceFilter[…,x,{y0,y-1,…}]

uses a specified list {y0,y-1,…} as the initial condition.

RecurrenceFilter[…,image]

filters image.

RecurrenceFilter[…,sound]

filters sampled sound object.

Details and Options
Details and Options Details and Options
Examples  
Basic Examples  
Scope  
Options  
Padding  
Applications  
Properties & Relations  
See Also
Related Guides
History
Cite this Page

RecurrenceFilter

RecurrenceFilter[{α,β},x]

filters x using a linear recurrence equation with coefficients α and β.

RecurrenceFilter[tf,x]

uses a discrete-time filter defined by the TransferFunctionModel tf.

RecurrenceFilter[…,x,{y0,y-1,…}]

uses a specified list {y0,y-1,…} as the initial condition.

RecurrenceFilter[…,image]

filters image.

RecurrenceFilter[…,sound]

filters sampled sound object.

Details and Options

  • RecurrenceFilter[{α,β},x] gives the response y to the causal input x by solving the recurrence equation for n from 1 to Length[x], where i is Length[α] and j is Length[β].
  • RecurrenceFilter[{α,β},x] uses an initial condition .
  • RecurrenceFilter[{α,β},x] pads the input x so that . The values can be changed with the Padding option. With Padding->None, the response y effectively starts at yj, and the output dimensions are normally smaller than the input. »
  • The specified coefficients α and β are respectively the denominator and numerator polynomial coefficients of the desired transfer function model.
  • In RecurrenceFilter[tf,…], tf should be a single-input single-output (SISO) system.
  • RecurrenceFilter works with arbitrary-rank numerical arrays, 2D and 3D images, and sampled sound objects, operating separately on each channel.
  • Possible sound objects include:
  • SampledSoundList[{a1,a2,…},r]amplitude levels given in a list
    SampledSoundFunction[f,n,r]amplitude levels generated by a function
    Sound[prims,…]excluding SoundNote objects in prims
  • When applied to images and multidimensional arrays, the specified filter is applied successively to each dimension, starting at level 1.
  • For multichannel image and sound objects, RecurrenceFilter is applied to each channel separately.

Examples

open all close all

Basic Examples  (2)

Filtering of a unit impulse sequence:

Wolfram Language code: RecurrenceFilter[{{1, -(1/2)}, {1}}, {1, 0, 0, 0, 0, 0}]

Filtering of an image:

Wolfram Language code: RecurrenceFilter[{{1, -(4/5)}, {(1/5)}}, [image]]

Scope  (5)

Use symbolic values for filter coefficients:

Wolfram Language code: RecurrenceFilter[{{1, Subscript[α, 1]}, {β}}, {1, 0, 0, 0, 0, 0}]

Filter a symbolic sequence:

Wolfram Language code: RecurrenceFilter[{{1, -(1/2)}, {1}}, {a, b, c, d}]

Apply a filter specified using a transfer function model:

Wolfram Language code: RecurrenceFilter[TransferFunctionModel[{{{z}}, -1/2 + z}, z, SamplingPeriod -> 1], {1, 0, 0, 0, 0, 0}]

Specify an initial condition:

Wolfram Language code: RecurrenceFilter[{{1, -(1/2)}, {1}}, {1, 0, 0, 0, 0, 0}, {1}]

Filter a TimeSeries:

Wolfram Language code: ts = TemporalData[TimeSeries, {{{0., -0.27267267057145633, -0.6672983789995302, -0.5338541947930846, -0.6117404489279314, -0.6755527076595494, -0.02125421294486496, -0.10792797291843935, -0.6138271235477938, -0.3248568606554575, -0.08843449054 ... 2053424, -0.49980440691873723, -0.5388679788215971, -0.4101602764645551}}, {{0, 1., 0.01}}, 1, {"Continuous", 1}, {"Continuous", 1}, 1, {ValueDimensions -> 1, ResamplingMethod -> {"Interpolation", InterpolationOrder -> 1}}}, False, 10.1];
Wolfram Language code: filtered = RecurrenceFilter[{{.3}, {.1}}, ts]
Wolfram Language code: ListLinePlot[{ts, filtered}, PlotLegends -> {"original data", "filtered"}]

Options  (1)

Padding  (1)

By default, RecurrenceFilter uses zero padding:

Wolfram Language code: RecurrenceFilter[{{1}, {1 / 3, 1 / 3, 1 / 3}}, Range[6]]

Use nonzero constant padding:

Wolfram Language code: RecurrenceFilter[{{1}, {1 / 3, 1 / 3, 1 / 3}}, Range[6], Padding -> -1]

Use periodic padding:

Wolfram Language code: RecurrenceFilter[{{1}, {1 / 3, 1 / 3, 1 / 3}}, Range[6], Padding -> "Periodic"]

With Padding->None, a shorter sequence is returned:

Wolfram Language code: RecurrenceFilter[{{1}, {1 / 3, 1 / 3, 1 / 3}}, Range[6], Padding -> None]

Applications  (3)

Filtering a noisy sinusoid signal with a recursive smoothing filter:

Wolfram Language code: data = Table[Sin[i ^ 2 + i] + RandomReal[{-.2, .3}], {i, 0, Pi, 0.01}]; {ListLinePlot[data], ListLinePlot[RecurrenceFilter[{{1, -(4/5)}, {(1/5)}}, data]]}

Filter an audio signal:

Wolfram Language code: s = Sound[SampledSoundList[Table[Cos[( i/4.) + ((i/50.))^2], {i, 8000}], 8000]]
Wolfram Language code: RecurrenceFilter[{{1, -(4/5)}, {(1/5)}}, s]

Filter an image using a lowpass Butterworth filter:

Wolfram Language code: RecurrenceFilter[ToDiscreteTimeModel[ButterworthFilterModel[{3, .5}, s], 1], [image]]

Properties & Relations  (5)

Impulse response of first-order recursive filter:

Wolfram Language code: RecurrenceFilter[{{1, -(1/2)}, {1}}, {1, 0, 0, 0, 0, 0}]

Step response of a non-recursive filter:

Wolfram Language code: RecurrenceFilter[{{1}, {1 / 3, 1 / 3, 1 / 3}}, {1, 1, 1, 1, 1, 1}]

In 2D, the filter is applied successively to rows and columns:

Wolfram Language code: data = (⁠| | | | | | | - | - | - | - | - | | 1 | 1 | 0 | 0 | 0 | | 1 | 1 | 0 | 0 | 0 | | 0 | 0 | 0 | 0 | 0 | | 0 | 0 | 0 | 0 | 0 | | 0 | 0 | 0 | 0 | 0 |⁠); RecurrenceFilter[{{1, -1 / 2}, {1}}, data]//MatrixForm

Use Map to apply the filter to rows only:

Wolfram Language code: Map[RecurrenceFilter[{{1, -1 / 2}, {1}}, #]&, data]//MatrixForm

For 1D arrays, OutputResponse is equivalent to RecurrenceFilter:

Wolfram Language code: sys = TransferFunctionModel[(z/z - (1/2)), z, SamplingPeriod -> 1]; OutputResponse[sys, {1, 0, 0, 0, 0, 0, 0, 0}][[1]] == RecurrenceFilter[sys, {1, 0, 0, 0, 0, 0, 0, 0}]

Use RecurrenceTable when describing the filter using a difference equation:

Wolfram Language code: RecurrenceTable[{y[n] - (1/2)y[n - 1] == DiscreteDelta[n], y[-1] == 0}, y[n], {n, 0, 7}] == RecurrenceFilter[{{1, -(1/2)}, {1}}, {1, 0, 0, 0, 0, 0, 0, 0}]

ListConvolve and RecurrenceFilter return equivalent results for non-recursive digital filters:

Wolfram Language code: ker = {1 / 3, 1 / 3, 1 / 3};data = {0, 0, 1, 1, 1, 1, 0, 0}; ListConvolve[ker, data] == RecurrenceFilter[{{1}, ker}, data][[3 ;; ]]
Wolfram Research (2012), RecurrenceFilter, Wolfram Language function, https://reference.wolfram.com/language/ref/RecurrenceFilter.html (updated 2014).

Text

Wolfram Research (2012), RecurrenceFilter, Wolfram Language function, https://reference.wolfram.com/language/ref/RecurrenceFilter.html (updated 2014).

CMS

Wolfram Language. 2012. "RecurrenceFilter." Wolfram Language & System Documentation Center. Wolfram Research. Last Modified 2014. https://reference.wolfram.com/language/ref/RecurrenceFilter.html.

APA

Wolfram Language. (2012). RecurrenceFilter. Wolfram Language & System Documentation Center. Retrieved from https://reference.wolfram.com/language/ref/RecurrenceFilter.html

BibTeX

@misc{reference.wolfram_2026_recurrencefilter, author="Wolfram Research", title="{RecurrenceFilter}", year="2014", howpublished="\url{https://reference.wolfram.com/language/ref/RecurrenceFilter.html}", note=[Accessed: 14-November-2026]}

BibLaTeX

@online{reference.wolfram_2026_recurrencefilter, organization={Wolfram Research}, title={RecurrenceFilter}, year={2014}, url={https://reference.wolfram.com/language/ref/RecurrenceFilter.html}, note=[Accessed: 14-November-2026]}