import numpy as np

MSD_list = [(0,0,0),(0.5,1,0),(1,2,0),(0,0,1),(2,1,1),(4,2,1)]
MSD_traces = [[] for x in range(2)]

#   MSD_traces is a list that collects the data later on in a nested loop.

MSD_array = np.zeros(len(MSD_list),dtype = [('d_space', 'f8'),('d_time', 'i4'), ('traceID', 'i4')])
MSD_array[:] = MSD_list

#   Now I want to go into the array, select all the data that has a common traceID,
#   then I select a subarray with same timelag and go into a second inner loop...

for i in range(2): #2 = the amount of unique traces in my example

    trace_view = MSD_array[MSD_array['traceID']==i]
    for k in range(np.max(trace_view['d_time'])+1):
        #iterating from d_time 0 to the maximum d_time in that trace

        time_view = trace_view[trace_view['d_time']==k]
        if time_view['d_space'].size:
            #checking for empty array where np.mean() yields error

            msd_mean = np.mean(time_view['d_space'])
            MSD_traces[i].append((msd_mean, k))

print(MSD_traces[0])
print(MSD_traces[1])