to_numeric() for safe conversion in Pandas - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of to_numeric() with errors='coerce'

What is the output of the following code snippet?

Pandas

import pandas as pd
s = pd.Series(['10', '20', 'abc', '30'])
result = pd.to_numeric(s, errors='coerce')
print(result)

A

0    10.0
1    20.0
2     NaN
3    30.0
dtype: float64

B

0     10
1     20
2    abc
3     30
dtype: object

C

0    10
1    20
2    NaN
3    30
dtype: float64

D ValueError: Unable to parse string 'abc' at position 2

💡 Hint

errors='coerce' replaces invalid parsing with NaN.

The to_numeric() function converts the series to numbers. With errors='coerce', invalid values like 'abc' become NaN. The result is a float series because of NaN.

❓ data_output

intermediate

Count of non-numeric values after conversion

After converting this Series with errors='coerce', how many values become NaN?

Pandas

import pandas as pd
s = pd.Series(['5', '7', 'nine', '11', 'twelve'])
converted = pd.to_numeric(s, errors='coerce')
count_nan = converted.isna().sum()
print(count_nan)

A 1

B 0

C 3

D 2

💡 Hint

Check which strings cannot be converted to numbers.

'nine' and 'twelve' are not numeric and become NaN, so count is 2.

🔧 Debug

advanced

Identify the error in to_numeric() usage

What error does this code raise?

Pandas

import pandas as pd
s = pd.Series(['1', '2', 'three'])
result = pd.to_numeric(s, errors='raise')
print(result)

A No error, output is a numeric series with NaN

B KeyError: 'three'

C ValueError: Unable to parse string 'three' at position 2

D TypeError: errors parameter must be 'ignore', 'coerce', or 'raise'

💡 Hint

errors='raise' stops on first invalid parsing.

With errors='raise', to_numeric() throws a ValueError when it finds a non-numeric string.

🚀 Application

advanced

Safe conversion of mixed data with fallback

You have a DataFrame column with mixed types. Which option safely converts the column to numeric, replacing errors with zero?

Pandas

import pandas as pd
df = pd.DataFrame({'values': ['100', 'abc', '200', None, '300']})

A df['values'] = pd.to_numeric(df['values'], errors='coerce').fillna(0)

B df['values'] = pd.to_numeric(df['values'], errors='ignore').fillna(0)

C df['values'] = pd.to_numeric(df['values'], errors='raise').fillna(0)

D df['values'] = df['values'].astype(float).fillna(0)

💡 Hint

Use errors='coerce' to convert invalid to NaN, then fillna to replace.

Option A converts invalid entries to NaN, then replaces NaN with 0. Option A leaves invalid as original strings, so fillna does nothing. Option A raises error on invalid. Option A fails because astype(float) raises error on 'abc'.

🧠 Conceptual

expert

Understanding dtype changes with to_numeric()

After applying pd.to_numeric() with errors='coerce' on a Series containing integers and invalid strings, what is the resulting dtype and why?

A int64, because original numbers are integers

B float64, because NaN requires a floating-point dtype

C object, because mixed types remain

D string, because invalid strings are kept

💡 Hint

NaN is a special floating-point value.

When invalid strings become NaN, the Series must use a floating-point dtype to hold NaN values, so dtype becomes float64.