fixed ---

I changed to a while loop and hard-reset the bin index after every full iteration of the inner loop. The code now does what it needs to do :D

LENGTH = len(rawdata)  # dimension of problem
      
bindex=0
dindex=0
while bindex<=2:
    while dindex<=LENGTH-1:
        if dfbins.loc[bindex][1] < dfrawdata.loc[dindex][0] < dfbins.loc[bindex][3]:
            emptyarray[bindex].append(dfrawdata.loc[dindex][1])  
        dindex = dindex + 1
    dindex = 0
    bindex = bindex +1  

This above block does it for 2 bins only because while bindex <=2

I can change that condition and it will do the whole data.

But it's still very slow.

For a problem where len(bins) = 450 and len(rawdata)=160000 it needs optimisation.

So the question is: how to more elegantly (pythonically) write a piece of python code that does the same as what I wrote?