Large data (multivariate polynomials) successfully written but unreadable by PARI-GP
Ewan Delanoy <[email protected]> Thu, 11 Dec 2025 17:57:37 +0100
| Newsgroups | gmane.comp.mathematics.pari.user |
|---|---|
| Message-ID | <[email protected]> |
In my current project, I'm dealing with polynomials in six variables and integer coefficients, max degree 20 for each variable (although the degrees are often smaller). A certain program I made successfully computes about 300.000 of those polynomials, and also successfully writes them to a script gp file. However, PARI-GP is unable to read the script even when I enlarge the PARI stack as much as I can, and exits suddenly with the laconic message "Killed" after trying. This is clearly a memory issue. The most obvious candidate for improvement in my program seems to be my very naive use of the bare "write" function to write those polynomials to disk. Some time ago, with the help of this mailing list I learnt how to improve the storage of univariate polynomials using Pol() and readvec. Are there similar optimizations for multivariate polynomials ?