CODEWITHSUNDEEP
pythonpandasnumpymatplotlib
Amar Shivaram
Amar Shivaram

Reputation: 153

TypeError: 'DataFrame' object is not callable

I've programmed these for calculating Variance

credit_card = pd.read_csv("default_of_credit_card_clients_Data.csv", skiprows=1)
    
for col in credit_card:
    var[col]=np.var(credit_card(col))

I'm getting this error

Traceback (most recent call last):
   File "C:/Python34/project.py", line 11, in <module>
     var[col]=np.var(credit_card(col)) 
TypeError: 'DataFrame' object is not callable

Upvotes: 14

Views: 243344

Answers (3)

Harry Moreno
Harry Moreno

Reputation: 11603

I got the same error when mistaking boolean indexing.

self_views = views[views('author_id') == views('viewer_id')]

the correct syntax is

self_views = views[views['author_id'] == views['viewer_id']]

Upvotes: 0

cottontail
cottontail

Reputation: 23021

This error occurs when you call a pandas DataFrame object - use round () brackets - as if it were a class or a function. Long story short, pandas DataFrames are objects of type 'DataFrame' whose attribute that makes an object callable is null.

For example, in the OP, the culprit is:

credit_card(col)

because previously, credit_card was defined to be a pandas DataFrame object via

credit_card = pd.read_csv("default_of_credit_card_clients_Data.csv", skiprows=1)

The most common way to define a dataframe is via the constructor df = pd.DataFrame(data) and this df cannot be called either: df() would give the same error. The list of other functions that construct a dataframe object can be found in the pandas documentation.

This problem is essentially a typo. Here are some common causes/solutions:

  1. Perhaps you were trying to select some columns of the dataframe, in which case, use the square [] brackets: 
    credit_card[col]                # e.g. credit_card['colA'] or credit_card[['colA']]
  2. Perhaps you were trying to call a dataframe method, in which case, write the method's name after the dataframe's name, separated by a dot .: 
    credit_card.some_method(col)    # e.g. credit_card.value_counts('colA')
  3. Perhaps you have a function and a dataframe sharing the same name in your environment. For example, 
    credit_card = lambda x: x + 1
credit_card = pd.DataFrame(data)
    In that case, make sure to re-define them using different names.

Upvotes: 4

jezrael
jezrael

Reputation: 862431

It seems you need DataFrame.var:

Normalized by N-1 by default. This can be changed using the ddof argument

var1 = credit_card.var()

Sample:

#random dataframe
np.random.seed(100)
credit_card = pd.DataFrame(np.random.randint(10, size=(5,5)), columns=list('ABCDE'))
print (credit_card)
   A  B  C  D  E
0  8  8  3  7  7
1  0  4  2  5  2
2  2  2  1  0  8
3  4  0  9  6  2
4  4  1  5  3  4

var1 = credit_card.var()
print (var1)
A     8.8
B    10.0
C    10.0
D     7.7
E     7.8
dtype: float64

var2 = credit_card.var(axis=1)
print (var2)
0     4.3
1     3.8
2     9.8
3    12.2
4     2.3
dtype: float64

If need numpy solutions with numpy.var:

print (np.var(credit_card.values, axis=0))
[ 7.04  8.    8.    6.16  6.24]

print (np.var(credit_card.values, axis=1))
[ 3.44  3.04  7.84  9.76  1.84]

Differences are because by default ddof=1 in pandas, but you can change it to 0:

var1 = credit_card.var(ddof=0)
print (var1)
A    7.04
B    8.00
C    8.00
D    6.16
E    6.24
dtype: float64

var2 = credit_card.var(ddof=0, axis=1)
print (var2)
0    3.44
1    3.04
2    7.84
3    9.76
4    1.84
dtype: float64

Upvotes: 10

Related Questions