How to Convert a String to Float in Python (Commas, Errors)

To convert a string to a float in Python, pass it to float(): float("19.99") returns 19.99. If the string contains thousands separators, remove them first (float("1,234.5".replace(",", ""))), and wrap the call in try/except ValueError when the text may not be a number. This guide shows how to convert string to float in Python in every common situation, and how to fix ValueError: could not convert string to float.

All examples were run with Python 3.12.5 and pandas 3.0.6 in the Windows Command Prompt. See the Python documentation for float() and locale.atof().

Convert a string to a float with float()

price = "19.99"
rate = float(price)                   # string -> float

print(rate, type(rate))
print(float("  42  "))                # spaces and newlines around the number are ignored
print(float("-3.5e2"))                # scientific notation works
print(float("1_000.5"))               # underscores between digits too (Python 3.6+)

Output:

19.99 <class 'float'>
42.0
-350.0
1000.5

float() accepts leading and trailing whitespace, a sign, a decimal point, an exponent (e or E), underscores between digits, and the words nan, inf and infinity. Anything else raises a ValueError.

ValueError: could not convert string to float

This error means the text contains something float() cannot read. These are the usual culprits:

values = ["3.14", "", "None", ".", "abc", "1,234.50", "12%", "$9.99"]

for v in values:
    try:
        print(f"{v!r:>12} -> {float(v)}")
    except ValueError as e:
        print(f"{v!r:>12} -> ValueError: {e}")

Output:

      '3.14' -> 3.14
          '' -> ValueError: could not convert string to float: ''
      'None' -> ValueError: could not convert string to float: 'None'
         '.' -> ValueError: could not convert string to float: '.'
       'abc' -> ValueError: could not convert string to float: 'abc'
  '1,234.50' -> ValueError: could not convert string to float: '1,234.50'
       '12%' -> ValueError: could not convert string to float: '12%'
     '$9.99' -> ValueError: could not convert string to float: '$9.99'

Error message ends with	Cause	Fix
''	An empty string (blank cell, empty input)	Check if text.strip(): or return a default
'None' / 'N/A'	A placeholder written as text	Map it to None or float("nan")
'1,234.50'	Thousands separator	text.replace(",", "")
'12%', '$9.99'	Symbols around the number	Remove them with re.sub() or strip("$%")
'.', 'abc'	Not a number at all	Validate or skip the value

Passing None itself (not the string) raises a different error: TypeError: float() argument must be a string or a real number, not 'NoneType'.

Convert a string with commas to float

Remove the thousands separator before converting. For European formats, where the comma is the decimal separator, swap the characters, or use the locale module:

import locale

us = "1,234,567.89"                                  # comma as thousands separator
print(float(us.replace(",", "")))

eu = "1.234.567,89"                                  # European: dot thousands, comma decimals
print(float(eu.replace(".", "").replace(",", ".")))

locale.setlocale(locale.LC_NUMERIC, "de_DE")         # or let the locale module do it
print(locale.atof(eu))
locale.setlocale(locale.LC_NUMERIC, "en_US")
print(locale.atof(us))

Output:

1234567.89
1234567.89
1234567.89
1234567.89

Locale names differ between systems ("de_DE" works on Windows and most Linux systems with that locale installed). setlocale() changes a global setting for the whole program, so prefer replace() inside libraries and web apps.

A safe string to float function

For messy data (prices, percentages, blanks), clean the text and return a default instead of crashing:

import re

def to_float(text, default=None):
    """Convert text such as '$1,299.50', ' 12 % ' or '' to a float, or return default."""
    if text is None:
        return default
    cleaned = re.sub(r"[^0-9eE.+-]", "", str(text))     # drop $, commas, %, spaces ...
    try:
        return float(cleaned)
    except ValueError:
        return default

for v in ["$1,299.50", " 12 % ", "", None, "N/A", "7.5e3", "-0.25"]:
    print(f"{v!r:>12} -> {to_float(v)}")

Output:

 '$1,299.50' -> 1299.5
    ' 12 % ' -> 12.0
          '' -> None
        None -> None
       'N/A' -> None
     '7.5e3' -> 7500.0
     '-0.25' -> -0.25

Convert user input to float

text = input("Enter a price: ")
try:
    price = float(text.replace(",", ""))
    print(f"Price with 18% tax: {price * 1.18:,.2f}")
except ValueError:
    print(f"'{text}' is not a number")

Output (the user typed 2,499.90):

Enter a price: 2,499.90
Price with 18% tax: 2,949.88

Convert a list of strings to floats

readings = ["21.5", "22.0", "", "23.4", "error", "24.1"]

numbers = [float(r) for r in readings if r.replace(".", "", 1).isdigit()]   # skip non-numbers
print(numbers)
print("average:", round(sum(numbers) / len(numbers), 2))

Output:

[21.5, 22.0, 23.4, 24.1]
average: 22.75

str.isdigit() alone rejects negative numbers and exponents; for those, use the to_float() function above in the list comprehension.

Convert a pandas column from string to float

pd.to_numeric(..., errors="coerce") converts a whole column and turns values it cannot parse into NaN, so one bad cell does not stop the conversion:

import pandas as pd

df = pd.DataFrame({"product": ["A", "B", "C", "D"],
                   "price": ["1,299.00", "849.50", "N/A", "2,050.75"]})

df["price_float"] = pd.to_numeric(df["price"].str.replace(",", ""), errors="coerce")   # bad values -> NaN
print(df)
print(df.dtypes)

Output:

  product     price  price_float
0       A  1,299.00      1299.00
1       B    849.50       849.50
2       C       N/A          NaN
3       D  2,050.75      2050.75
product            str
price              str
price_float    float64
dtype: object

When reading files, pd.read_csv("file.csv", thousands=",") (and decimal="," for European files) converts the numbers while loading.

Check if a string can be converted to float

def is_float(text):
    try:
        float(text)
        return True
    except (TypeError, ValueError):
        return False

for v in ["3.14", "10", "1e5", "nan", "abc", "", "1,5"]:
    print(f"{v!r:>7} {is_float(v)}")

Output:

 '3.14' True
   '10' True
  '1e5' True
  'nan' True
  'abc' False
     '' False
  '1,5' False

Note that "nan" counts as a valid float. For money, consider decimal.Decimal(text) instead of float, because binary floats cannot store values like 0.1 exactly.

Related Python conversion tutorials:

• Convert a dictionary to a string in Python
• Convert datetime to string in Python
• Remove characters from a string in Python
• Get the first and last digit of a number

Frequently asked questions