BUG, ENH: fix array2string rounding bug by adding min_digits option

[ahaldane]

Fixes #18609

This PR adds a new min_digits argument to the dragon4 float printing methods, such as np.format_float_positional. This kwd guarantees that at least the given number of digits will be printed, when printing in unique=True mode, even if the extra digits are unnecessary to uniquely specify the value. It is the counterpart to the precision argument which sets the maximum number of digits to be printed. (When unique=False fixed-precision mode, it has no effect, and the precision arg fixes the number of digits)

Using this new keyword, I then fix a bug in arrayprint ( #18609). The bug in the old code is that, in order to guarantee that all elements of an array print using the same number of digits, we turned off unique=True because we needed to use fixed-precision mode to fix the number of digits. However, unique mode rounds differently from fixed-precision mode, so the rounding was wrong. This PR instead now stays in unique mode to print the elements, but uses the min_digits and precision arguments to fix the number of output digits.

It looks like a lof of chages, but a lot of it is just introducing the min_digits arg into all the function signatures in dragon4 and in the user api.

[BvB93]

Can you update the annotations in the arrayprint.pyi stub file?
By the looks of it the type of min_digits should be Optional[int].

numpy/numpy/core/arrayprint.pyi

Lines 98 to 116 in ee9c8f8

	def format_float_scientific(
	x: _FloatLike_co,
	precision: Optional[int] = ...,
	unique: bool = ...,
	trim: Literal["k", ".", "0", "-"] = ...,
	sign: bool = ...,
	pad_left: Optional[int] = ...,
	exp_digits: Optional[int] = ...,
	) -> str: ...
	def format_float_positional(
	x: _FloatLike_co,
	precision: Optional[int] = ...,
	unique: bool = ...,
	fractional: bool = ...,
	trim: Literal["k", ".", "0", "-"] = ...,
	sign: bool = ...,
	pad_left: Optional[int] = ...,
	pad_right: Optional[int] = ...,
	) -> str: ...

[seberg]

Thanks Allan. I took the liberty to rebase, if we want a backport give me a ping and I will do that.

I won't claim the have followed all the logic in dragon4.c but the code changes look all good and the tests sufficiently thorough.

[charris]

Could use a release note.

[charris]

Thanks Allan