Basically, you enter the same keystrokes as you would with an RPN calculator (such as an HP 48). One difference is that you separate numbers, operators, and functions by a space or comma. The RPN calculator in OpenDSS has a "stack" of 10 registers just like some calculators. In the function descriptions that follow, the first stack register is referred to as "X" and the next higher as "Y" as on the HP calculator. When a new number is entered, the existing registers are rolled up and the new number is inserted into the X register.

These functions operate on the first two registers, X and Y:

+        Add the last two operands (X and Y registers; result in X; X = X + Y )

-        X = Y - X

*        X = X * Y

/        X = Y / X

^        (exponentiation) X = Y to the X power

These unitary functions operate on the X register and leave the result in X. 

sqr    X=X*X

sqrt   take the square root of X

inv    (inverse of X = 1/X)

ln     (natural log of X)

exp    (e to the X)

log10    (Log base 10 of X)

sin         for X in degrees, take the sine

cos         for X in degrees, take the cosine

tan         for X in degrees, take the tangent
asin       take the inverse sine, result in degrees

acos       take the inverse cosine, result in degrees

atan       take the arc tangent, result in degrees
atan2  take the arc tangent with two arguments, Y=rise and X = run, result in degrees over all four quadrants

The following functions manipulate the stack of registers

Rollup     shift all registers up

Rolldn     shift all registers dn

Swap          swap X and Y

There is one constant preprogrammed: pi