Understanding Slash and Subsequent Meaning in the Numpy Reference Function Arguments

Almost everything is explained in the link in the comment, but I will leave a reply in Japanese for your reference.

This notation is not unique to numpy, but is traditionally used in Python documentation.Originally, this was a notation to describe the unique specifications of Python functions implemented in C, but since Python 3.8 can now be used with pure Python code, I will explain based on that.

In the middle of an argument, / is a marker that divides the arguments of a function definition to the left and right, and does not actually pass the argument /.In fact, there has always been a delimited mark * on something similar.Roughly speaking, it means:

The arguments to the left of
• / are positional-only parameters and must be passed based on the order of the arguments when calling a function, not on keywords (like f(x=42) using the name of the provisional argument.
Arguments to the right of
• * are keyword-only parameters and must be passed with keywords.
Arguments between
• / and * can be specified by location or keyword (typically this is the default).

PEP570 for a specific example.

def standard_arg(arg):
    print(arg)

def pos_only_arg(arg, /):
    print(arg)

defkwd_only_arg(*,arg):
    print(arg)

def combined_example(pos_only, /, standard, *, kwd_only):
    print(pos_only, standard, kwd_only)

When there is such a function, the first function can be called either standard_arg(2) or standard_arg(arg=2), but the second function is an error if pos_only_arg(arg=1).Similarly, if the third function is kwd_only_arg(3), it will fail.The last example is a mixed pattern, and the relative position between / and * determines how the arguments are passed.

For more information, see the PEP570 or the Python 3.8 Release Notes location-only argument item .