Hey there! Let's break down your two questions about controlling float output in Common Lisp—first the general case, then the Antik + SLY specific scenario.

1. General Common Lisp: Reducing Displayed Digits for Double-Floats #

The go-to tool for formatting float output in Common Lisp is the format function. It lets you control exactly how many decimal places are shown, without altering the underlying double-precision value itself.

• To print directly to the console (or REPL), use:

  (format t "~,4F" (exp 1d0)) ; Outputs 2.7183
  

  Here's what the format specifier means:

  • ~ starts a format directive
  • ,4 specifies we want 4 decimal places
  • F denotes fixed-point floating-point formatting

• If you need the formatted result as a string (instead of printing it), replace t with nil:

  (format nil "~,4F" (exp 1d0)) ; Returns the string "2.7183"
  

This approach keeps the original double-float value intact—you're only changing how it's displayed.

2. Antik + Emacs/SLY: Reducing Output While Preserving Calculation Precision #

Since you've already set *read-default-float-format* to 'double-float (which ensures inputs are parsed as double-precision), here are two reliable ways to shorten output without losing calculation accuracy:

Option 1: Use format with Antik results #

Just like the general case, wrap Antik function calls in format to control display:

;; Example with an Antik function that returns a double-float
(format t "~,4F" (antik:some-antik-function)) ; Shows 4 decimal places

Option 2: Adjust global REPL output settings #

If you want all float outputs in your SLY REPL to show fewer digits by default, set the *print-float-digits* variable. This affects how the REPL displays all floats, while calculations still use full double-precision:

(setf *print-float-digits* 4)

Now when you evaluate (exp 1d0) in the SLY REPL, it will output 2.7183 instead of the full precision string. Note this is a global setting—all subsequent float outputs in the REPL will follow this until you change it back.

A quick reminder: *read-default-float-format* only affects how the Lisp reader interprets input floats (making sure plain 1.0 is read as a double-float instead of single-float). It doesn't control output—format or *print-float-digits* handle that part.

内容的提问来源于stack exchange，提问作者dchagas