Finding elements in string or cell array with common strings

Consider two arrays A & B:

A={'blah_12_blah' 'blah_456_blah' 'blah_789_blah' 'blah_NPQZ_blah'}
B={'blah_NPQZ_blah' 'blah_135_blah' 'blah_579_blah' 'blah_12_blah' 'blah_RSTX_blah'} 

The termn 'blah' basically refers to any string that comes before and after the parts of interest and each 'blah' could be a different string.

I would like to compare these two arrays and fine only the members that have the "part of interest" in common. In this case it means that after comparing A & B, only the elements 'blah_12_blah' and 'blah_NPQZ_blah' will show up as output, since these elements have the parts '12' and 'NPQZ' incommon, no matter what their respective 'blah' parts are.

Answer 1

[Kevin Holly]

A={'blah_12_blah' 'blah_456_blah' 'blah_789_blah' 'blah_NPQZ_blah'}

A = 1×4 cell array
    {'blah_12_blah'}    {'blah_456_blah'}    {'blah_789_blah'}    {'blah_NPQZ_blah'}

B={'blah_NPQZ_blah' 'blah_135_blah' 'blah_579_blah' 'blah_12_blah' 'blah_RSTX_blah'} 

B = 1×5 cell array
    {'blah_NPQZ_blah'}    {'blah_135_blah'}    {'blah_579_blah'}    {'blah_12_blah'}    {'blah_RSTX_blah'}

A_partofinterest = extractBetween(A,'_','_')

A_partofinterest = 1×4 cell array
    {'12'}    {'456'}    {'789'}    {'NPQZ'}

B_partofinterest = extractBetween(B,'_','_')

B_partofinterest = 1×5 cell array
    {'NPQZ'}    {'135'}    {'579'}    {'12'}    {'RSTX'}

A_new = A(ismember(A_partofinterest,B_partofinterest))

A_new = 1×2 cell array
    {'blah_12_blah'}    {'blah_NPQZ_blah'}

B_new = B(ismember(B_partofinterest,A_partofinterest))

B_new = 1×2 cell array
    {'blah_NPQZ_blah'}    {'blah_12_blah'}

[Saeid]

Hi Kevin, this is really good, and I didn't know that this command existed.

However, it looks like I misrepresented the question by adding the '_' character implying that it is a fixed part of the member. Actually, it is not, and I just put it there to separate the blahs from the parts of interest for better visibility. In other words, the elements would look more like this:

'blah123blah'
'blah456blah' 

etc.

[Kevin Holly]

That makes it a bit more challenging. Here is my solution.

A={'blah12balh' '54324569689' 'asdfkj789oiups' '+_()_NPQZ@$#%'}

A = 1×4 cell array
    {'blah12balh'}    {'54324569689'}    {'asdfkj789oiups'}    {'+_()_NPQZ@$#%'}

B={'wwelnkNPQZcvou' 'utcim135xzus' 'rsnfu579sodm' '9f6ej12d9j37' '0o1q9i2wRSTX7y3e64tr'} 

B = 1×5 cell array
    {'wwelnkNPQZcvou'}    {'utcim135xzus'}    {'rsnfu579sodm'}    {'9f6ej12d9j37'}    {'0o1q9i2wRSTX7y3e64tr'}

Assuming two characters have to be the same.

A_new_index = [];
B_new_index = [];
for b = 1:length(B)
    for a = 1:length(A)
        for ii = 1:length(A{a})-1
            if contains(B{b},A{a}(ii:ii+1)) == 1
                display(['part of interest is ' A{a}(ii:ii+1)])
                display(['A cell is ' A{a}])
                A_new_index = [A_new_index a];
                display(['B cell is ' B{b}])
                B_new_index = [B_new_index b];
            end
        end
    end
end

part of interest is NP

A cell is +_()_NPQZ@$#%

B cell is wwelnkNPQZcvou

part of interest is PQ

A cell is +_()_NPQZ@$#%

B cell is wwelnkNPQZcvou

part of interest is QZ

A cell is +_()_NPQZ@$#%

B cell is wwelnkNPQZcvou

part of interest is 12

A cell is blah12balh

B cell is 9f6ej12d9j37

A_new = A(1,unique(A_new_index))

A_new = 1×2 cell array
    {'blah12balh'}    {'+_()_NPQZ@$#%'}

B_new = B(1,unique(B_new_index))

B_new = 1×2 cell array
    {'wwelnkNPQZcvou'}    {'9f6ej12d9j37'}

[Saeid]

Thanks Kevin, I guess this will work, but just one final question about a problem that is somewhere between the first and the second one: what if the reference array (A) has fixed elements (no blahs) and the second one is as before?

A=[123abc 456klm 1a2b3c ...]
B=[blah123abcblah blah789ijkblah blah11aa22bbblah 123abc 456pqr 1a2b3c ...]

In this case I want to find any element of B that has exactly 123abc somewhere in it and so on.

[Kevin Holly]

A={'123abc' '456klm' '1a2b3c'}

A = 1×3 cell array
    {'123abc'}    {'456klm'}    {'1a2b3c'}

B={'blah123abcblah' 'blah789ijkblah' 'blah11aa22bbblah' '123abc' '456pqr' '1a2b3c'}

B = 1×6 cell array
    {'blah123abcblah'}    {'blah789ijkblah'}    {'blah11aa22bbblah'}    {'123abc'}    {'456pqr'}    {'1a2b3c'}

contains(B,A)

ans = 1×6 logical array
   1   0   0   1   0   1

B(contains(B,A))

ans = 1×3 cell array
    {'blah123abcblah'}    {'123abc'}    {'1a2b3c'}

[Saeid]

Cool, thanks!