@msz59, it seems that your earlier comment about the strangeness of the values was correct.
Some of the values I expected for VArPolar[j][i+1000] were:
VArPolar[2900][1000] = 103;
VArPolar[3000][1000] = 64;
VArPolar[3000][1000] = 38.8;
Whereas, using this solution, assuming the same endiannes I obtain:
VArPolar[2900][1000] = -9.2819E-168;
VArPolar[3000][1000] = -3.19823E-295;
VArPolar[3000][1000] = -1.60563E-267;
Assuming different endiannes, I obtain:
VArPolar[2900][1000] = -4.35932E260;
VArPolar[3000][1000] = -1.49665E-192;
VArPolar[3000][1000] = 1.59033E75;
The data is supposed to be double-precision and so each value should be 8 bytes. But there must a mistake somwhere since I expect:
-10 < (VArPolar[j => 2900][1000]) < 110.
Do you have any further suggestions?