assign 0 when value_count() is not found-CodePudding

I have a column that looks like this:

group
A
A
A
B
B
C

The value C exists sometimes but not always. This works fine when the C is present. However, if C does not occur in the column, it throws a key error.

    value_counts = df.group.value_counts()
    new_df["C"] = value_counts.C

I want to check whether C has a count or not. If not, I want to assign new_df["C"] a value of 0. I tried this but i still get a keyerror. What else can I try?

    value_counts = df.group.value_counts()
    new_df["C"] = value_counts.C
    if (df.group.value_counts()['consents']):
      new_df["C"] = value_counts.consents
    else:
      new_df["C"] = 0

CodePudding user response：

One way of doing it is by converting series into dictionary and getting the key, unless not found return the default value (in your case it is 0):

df = pd.DataFrame({'group': ['A', 'A', 'B', 'B', 'D']})

new_df = {}

character = "C"
new_df[character] = df.group.value_counts().to_dict().get(character, 0)

output of new_df

{'C': 0}

However, I am not sure what new_df should be, it seems that it is a dictionary? Or it might be a new dataframe object?

CodePudding user response：

One way could be to convert the group column to Categorical type with specified categories. eg:

df = pd.DataFrame({'group': ['A', 'A', 'A', 'B', 'B']})
print(df)

#   group
# 0     A
# 1     A
# 2     A
# 3     B
# 4     B

categories = ['A', 'B', 'C']

df['group'] = pd.Categorical(df['group'], categories=categories)
df['group'].value_counts()

[out]

A    3
B    2
C    0
Name: group, dtype: int64