FIR_Filter

This reposite implements a FIR filter designing for lowpass, highpass, bandpass, bandstop. It is designed in C++. However, it is possible to use this FIR filter as a python module. Therefore, the open-source software SWIG is used to convert C/C++ language in script language as Python or Java. The code, a demo and an howto can be found on https://github.com/nathanLoretan/FIR_Filter.

Howto

To ease the compilation of the C++ class, a simple Makefile is created.

$ cd /path/to/FIR_filter
$ make

By default, the Makefile compile using python3 but it is possible to specify the version python2.7 or python3. It is important to clean the project if the python version is changed.

$ make python3.5
$ make python2.7

The makefile compiles the C++ class and create the python module. Then the python package can be simply improted. The FIR filter class is not compiled in the reposite.

Demo

In the FIR_filter_demo.py, four FIR filters are created of each possible type. A window function of hamming is used to improve those filters. The filters are characterized by:

• Sampling rate: 1kHz
• Number of taps: 800
• Low pass filter cut frequency: 50Hz
• High pass filter cut frequency: 50Hz
• Band pass filter cut frequencies: 50Hz - 100Hz
• Stop band filter cut frequencies: 50Hz - 100Hz

Then, a signal composed with three frequencies, 10Hz, 75Hz and 150Hz, is created and is filtered with the different types of FIR filter.

API

FIR_filter( int taps=0, double f1=0, double f2=0, char* type="", char* window="")

Create an FIR_filter object.

Parameters

• [in] taps : number taps used by the filter
• [in] f1 : 1st cut frequency normalized
• [in] f2 : 2st cut frequency normalized, used only with bandstop and passband
• [in] type : type of filter (lowpass, highpass, bandpass, stopband)
• [in] window : window function possible (hamming, hanning, triangle, blackman)

---

std::vector<double> getCoefficients()

Get the coefficients calculated for the specific filter.

Return

The coefficients used by the FIR filter.

---

double filter(double new_sample)

Filter the new sample given to the FIR filter.

Return

Result of the filtering

Parameters

• [in] new_sample : new sample