Wolfram Function Repository
Instant-use add-on functions for the Wolfram Language
Function Repository Resource:
Create a fast Fourier transform calculation in graphical form
ResourceFunction["FastFourierGraph"][dim] returns an edge-weighted, time-ordered multiway graph on 2dim×2dim vertices, which describes the fast Fourier transform of a length 2dim vector. | |
ResourceFunction["FastFourierGraph"][data] takes a list of data with length 2dim and adds it to the graph under the annotation key VertexWeight. |
Calculate the graph for the base case of the fast Fourier transform:
In[1]:= |
Out[1]= |
Plot the next induced case:
In[2]:= |
Out[2]= |
Color edges according to phases:
In[3]:= |
Out[3]= |
Notice n log n complexity in rectangular dimensions of subsequent graphs:
In[4]:= |
Out[4]= |
Blow up the graphs to see their free-form structure using "SpringElectricalEmbedding":
In[5]:= |
Out[5]= |
Create a graph from data:
In[6]:= |
Out[6]= |
Use the option "ComputeValues"→True to compute the transformation values while generating the graph:
In[7]:= |
Out[7]= |
See the fast Fourier transformation values by retrieving the VertexWeight:
In[8]:= |
Out[8]= |
Depict results of the calculation as above:
In[9]:= |
Out[9]= |
Compare with the standard method using FourierMatrix:
In[10]:= |
Out[10]= |
Apply a similar validation technique for different input lengths:
In[11]:= |
Out[11]= |
Input should be a list of length 2n for some integer n>0:
In[13]:= |
Out[13]= |
In[14]:= |
Out[14]= |
Timing statistics look very slow compared to the built-in Fourier:
In[15]:= |
Out[15]= |
In[16]:= |
Out[16]= |
Compile time is relatively slow compared to ButterflyGraph:
In[17]:= |
Out[17]= |
Follow the different paths a value can take from input to output:
In[18]:= |
Out[18]= |
Plot an intensity profile:
In[19]:= |
Out[19]= |
This work is licensed under a Creative Commons Attribution 4.0 International License