Why i-1 index for csum used in sum calculation

In the video,

for calculation of sum

i.e. sum=csum[j]-csum[i-1];

for i=0 and j=i (j=0)

what will be csum[i-1] ? (csum[-1])