"...one is the result of a foor loop using angle as the index (alpha = 25:89)"
That is the start of your problems: confusing data with code. This approach is fragile and inefficient. Any negative, zero, or fractional values will throw an error. Any "missing" angles (up to the maximum) will still have stored values in the data array, but serve no purpose.
A much more robust approach would be to store the angles in one vector and the results in another. Then all of your data vectors will have exactly the same size.
