On Thu, Sep 24, 2020 at 12:16 PM 'Pascal Jasmin' via Source < [email protected]> wrote:
> The source of the bytes can still be files. Just a separate step. As a > library, both options can be available. > Two-step construction to get J objects would make the facility unable to use for us, in terms of performance and memory cost. There is no existing npy reader and write that allows you operating the npy binary content in strings (you can use a memory file like StringIO, though). -- Zhihao ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
