Getting differnet answers fromMatlab for what looks like the same expression

I am trying to raise a real number (or vector of real numbers) to a fractional power.

V = -10:0.1:80;

V0 = 8;

m = 0.323;

Vx = (1 + V/V0);

Vx(1)

ans =

-0.25

>> sprintf('%0.55f', Vx(1))

ans =

'-0.2500000000000000000000000000000000000000000000000000000'

>> (Vx(1))^.323

ans =

0.337319506343289 + 0.542770712359585i

sprintf('%0.55f', -0.25)

ans =

'-0.2500000000000000000000000000000000000000000000000000000'

>> -0.25^.323

ans =

-0.6390

Why is the power of one version of 0.25 real, and the other is complex, when both versions of "-0.25" are identical to 55 sf?

Answer 1

[Fangjun Jiang]

Try this and hope you know why

(-0.25)^.323

[John Wood]

OK, got it.

[Steven Lord]

To be more explicit, the power and matrix power operators (.^ and ^ respectively) have higher precedence (level 2) than unary minus (the unary - operator) has (level 4.) See this page for the precedence breakdown. Parentheses are at level 1, which is why Fangjun Jiang's suggestion gives a complex result.