Agena

agena >> `The Power of Procedural Programming` 7.5.1 Tethys, April 28, 2026 - You can now change the precision of an existing MPFR value, real or complex, with `mpfr.precision`. Just pass the MPFR value as a first argument and the new precision in bits as a second. - You can also now pass the precision to `mpfr.clone`, as an optional second argument. - `linalg` vectors and matrices can now store numeric MAPM and MPFR numbers, as well as double-to-double values. The main underlying functions `linalg.setvelem` and `linalg.getvelem` have been extended accordingly and have been documented in the Agena Reference for the first time. - New `unimath.linsolve` solves a linear system of equations using Gaussian elimination with partial pivoting, supporting standard Agena numbers, double-to-double numbers, long doubles, numeric MAPM and MPFR values. - Likewise new `unimath.det` computes the determinant, also supporting standard Agena numbers, double-to-double numbers, long doubles, numeric MAPM and MPFR values. - `dd.renorm` did not work as described in the Agena Reference and has been fixed. You can either pass the high and the low part as standard Agena numbers or a double-to-double value. If you pass a double-to-double value and the Boolean value `true`, then the function returns a normalised double-to-double value, with the high and the low part in the input summed up and written to the high part and the low part set to zero. - Added the `telex` package which encodes and decodes a string to/from its 5-bit International Telegraphy Alphabet representation, packing it to around 70 % of its size. See Chapter 16.2 for more information.