Understanding this class in python. The operator % and formatting a float.

Question

class FormatFloat(FormatFormatStr):
    def __init__(self, precision=4, scale=1.):
        FormatFormatStr.__init__(self, '%%1.%df'%precision)
        self.precision = precision
        self.scale = scale

    def toval(self, x):
        if x is not None:
            x = x * self.scale
        return x

    def fromstr(self, s):
        return float(s)/self.scale

The part that confuses me is this part

FormatFormatStr.__init__(self, '%%1.%df'%precision)

does this mean that the precision gets entered twice before the 1 and once before df? Does df stand for anything that you know of? I don't see it elsewhere even in its ancestors as can be seen here:

class FormatFormatStr(FormatObj):
    def __init__(self, fmt):
        self.fmt = fmt

    def tostr(self, x):
        if x is None: return 'None'
        return self.fmt%self.toval(x)

class FormatObj:
    def tostr(self, x):
        return self.toval(x)

    def toval(self, x):
        return str(x)

    def fromstr(self, s):
        return s

also, I put this into my Ipython and get this:

In [53]: x = FormatFloat(.234324234325435)
In [54]: x
Out[54]: <matplotlib.mlab.FormatFloat instance at 0x939d4ec>

I figured that it would reduce precision to 4 and scale to 1. But instead it gets stored somewhere in my memory. Can I retrieve it to see what it does to the number?

Thanks everyone you're very helpful!

Answer 1

In ('%%1.%df' % precision), the first %% yields a literal %, %d is substituted with precision, and f is inserted literally. Here's an example of how it might turn out:

>>> '%%1.%df' % 4
'%1.4f'

More about string formatting in Python

In order to use the FormatFloat class you might try something like this:

formatter = FormatFloat(precision = 4)
print formatter.tostr(0.2345678)

Answer 2

In python format strings, "%%" means "insert a literal percent sign" -- the first % 'escapes' the second, in the jargon. The string in question, "%%1.%df" % precision is using a format string to generate a format string, and the only thing that gets substituted is the "%d". Try it at the interactive prompt:

>>> print "%%1.%df" % 5
'%1.5f'

The class FormatFloat doesn't define __repr__, __str__, or __unicode__ (the "magic" methods used for type coercion) so when you just print the value in an interactive console, you get the standard representation of instances. To get the string value, you would call the tostr() method (defined on the parent class):

>>> ff = FormatFloat(.234324234325435)
>>> ff.tostr()
0.234

Answer 3

>>> precision=4
>>> '%%1.%df'%precision
'%1.4f'

%% gets translated to %

1 is printed as is

%d prints precision as a decimal number

f is printed literally