Search code examples
pythonpandasnumpyadjacency-matrix

Create adjacency matrix from array and its transpose

I have a pandas DataFrame like this:
enter image description here

And I would like to create a symmetrical matrix like this:
enter image description here

Where the value is actually the length of the intersection of the two lists. Therefore I have made this function:

 def intersectSize(l1, l2):
     return len(set(l1) & set(l2))

Is there a function out there that would resemble this one:

def createSymMatrix(array, func):
    ...
    return matrix

where array is my initial dataframe and func is the intersectSize function?

EDIT: figured it out with this two liner: 

array = [[len(set(l1)&set(l2)) for l1 in df]] for l2 in df]
adj = pd.DataFrame(data=array, index=df.index, columns=df.index)
Solution

  • I think you need this,

    r=[]
for val in list(itertools.product(df[0].values,df[0].values)):
    r.append( len(set(val[0])&set(val[1])) )
print pd.DataFrame(np.array(r).reshape(len(df),-1))

    Using List Comprehension:

    t= [len(set(val[0])&set(val[1])) for val in list(itertools.product(df[0].values,df[0].values))]
print pd.DataFrame(np.array(t).reshape(len(df),-1))

    Output:

       0  1  2  3
0  3  0  0  1
1  0  1  0  1
2  0  0  2  0
3  1  1  0  2