I'm trying to handle a dataset containing only str, but the values of the str can be either real text or something like "1.0" and "3.54".
I want to convert all string to the best possible type, which means 1.0 should be converted to int, 3.54 should be converted to float and everything else should stay str.
What is the best way to do so?
lets try
def convert(txt): try: k = float(txt) if k%1 ==0: return int(k) return k except ValueError: return txt now what I'm thinking is, the input is either a number or not, anyway we can float it. if it is also divisable by 1 its an int and we're done. if it isn't then it is a float after all and we're done. any other case: there's nothing that we can do, and then let's return it
You can use a try except block to perform the operation.
for s in string: try: //convert to int or float except: pass 
int or float but those specifics is what the user is asking for.You may use
strings = ["some weird junk", "1.0", "2.345", -1000] for string in strings: try: number, decimal = string.split(".") try: if decimal == "0": value = int(number) else: value = float(string) except: value = string except (ValueError, AttributeError): try: value = int(string) except ValueError: value = string print(value, type(value)) Which yields
some weird junk <class 'str'> 1 <class 'int'> 2.345 <class 'float'> -1000 <class 'int'> 
def convert(string): possible = [str, int, float] for func in possible: try: result = func(string) except ValueError: continue if str(result) == string and type(result) is not str: return result return string strings = ['hello', '1', '1.0', '1.9'] for string in strings: result = convert(string) print(type(result), result) Output: <class 'str'> hello <class 'int'> 1 <class 'float'> 1.0 <class 'float'> 1.9 
"1.0"is clearly meant to represent a floating point number (even though one with an integer value) - why would you not want to convert it to float?