Cannot perform 'rand_' with a dtyped [int64] array and scalar of type [bool]

# Cannot perform 'rand_' with a dtyped [int64] array and scalar of type [bool]

The Pandas "TypeError: Cannot perform 'rand_' with a dtyped [int64] array and scalar of type [bool]" occurs when you try to filter a DataFrame based on multiple conditions without wrapping each condition in parentheses.

To solve the error, wrap the conditions in parentheses so that the order of precedence is correct.

Here is an example of how the error occurs.

main.py

Copied!
import pandas as pd

df = pd.DataFrame({
    'A': ['Bobby', 'Bobby', 'Carl', 'Dan'],
    'B': [175.1, 180.2, 190.3, 205.4],
    'C': [10, 15, 20, 25],
})

# ⛔️ TypeError: Cannot perform 'rand_' with a dtyped [int64] array and scalar of type [bool]
print(df.loc[df.A == 'Bobby' & df.C > 10])

type error cannot perform rand with dtyped array and scalar

We tried to filter the DataFrame based on multiple conditions but forgot to wrap each condition in parentheses ().

# Wrap each condition in parentheses to correct the order of precedence

When using the logical AND (&) operator to chain multiple conditions, make sure each condition is wrapped in parentheses so that the order of precedence in the expression is correct.

main.py

Copied!
import pandas as pd

df = pd.DataFrame({
    'A': ['Bobby', 'Bobby', 'Carl', 'Dan'],
    'B': [175.1, 180.2, 190.3, 205.4],
    'C': [10, 15, 20, 25],
})

#        A      B   C
# 1  Bobby  180.2  15
print(df.loc[(df.A == 'Bobby') & (df.C > 10)])

wrap each condition in parentheses

The code for this article is available on GitHub

The code sample returns a subset of the DataFrame where the A column is equal to the string "Bobby" and the C column is greater than 10.

Both conditions have to be met for the result to get added to the new DataFrame.

If you try to filter a DataFrame based on multiple "OR" conditions with a pipe | without wrapping each condition in parentheses, you would get an error.

main.py

Copied!
import pandas as pd

df = pd.DataFrame({
    'A': ['Bobby', 'Bobby', 'Carl', 'Dan'],
    'B': [175.1, 180.2, 190.3, 205.4],
    'C': [10, 15, 20, 25],
})

# ⛔️ TypeError: Cannot perform 'ror_' with a dtyped [int64] array and scalar of type [bool]
df2 = df.loc[df.A == 'Bobby' | df.C > 10]

As shown in the previous example, we have to wrap each condition in parentheses, so that the order of precedence is correct.

main.py

Copied!
import pandas as pd

df = pd.DataFrame({
    'A': ['Bobby', 'Bobby', 'Carl', 'Dan'],
    'B': [175.1, 180.2, 190.3, 205.4],
    'C': [10, 15, 20, 25],
})

df2 = df.loc[(df.A == 'Bobby') | (df.C > 10)]

#        A      B   C
# 0  Bobby  175.1  10
# 1  Bobby  180.2  15
# 2   Carl  190.3  20
# 3    Dan  205.4  25
print(df2)

wrap in parentheses when using or operator

The code for this article is available on GitHub

We used the logical OR | operator, so either condition has to be met.

For a row to get added to the resulting DataFrame:

Accessing the A column has to return the string "Bobby".
Or accessing the C column has to return a value greater than 10.

You also have to wrap each condition in parentheses when filtering a DataFrame in an assignment.

main.py

Copied!
import pandas as pd

df = pd.DataFrame({
    'A': ['Bobby', 'Bobby', 'Carl', 'Dan'],
    'B': [175.1, 180.2, 190.3, 205.4],
    'C': [10, 15, 20, 25],
})

df.loc[(df.A == 'Bobby') & (df.C < 14), 'C'] = 100000

print(df)

Running the code sample produces the following output.

shell

Copied!
       A      B       C
0  Bobby  175.1  100000
1  Bobby  180.2      15
2   Carl  190.3      20
3    Dan  205.4      25

wrap each condition in parentheses in assignment

The code for this article is available on GitHub

The A column has to have a value of "Bobby" and the C column has to have a value of less than 14 for the value of the C column to get updated to 100000.

Wrapping each condition in parentheses is documented in this section of the Pandas docs.

Wrapping the conditions in parentheses is necessary because:

By default, Python evaluates an expression such as df['A'] > 5 & df['B'] < 10 as df['A'] > (5 & df['B']) < 10.
Whereas, the correct evaluation order is (df['A'] > 5) & (df['B'] < 10).

The order of precedence of the & is set by Python and not by the Pandas library.

However, wrapping each condition in parentheses indicates to the language that the expressions between the parentheses have to be evaluated before the & operator.

# Additional Resources

You can learn more about the related topics by checking out the following tutorials:

I wrote a book in which I share everything I know about how to become a better, more efficient programmer.

You can use the search field on my Home Page to filter through all of my articles.

Cannot perform 'rand_' with a dtyped [int64] array and scalar of type [bool]