How to check if pandas Series is empty?

Question

How to check if pandas Series is empty?

I have tried this:

How to check whether a pandas DataFrame is empty?

but it seems that Series has no property 'isempty'.

Answer 1

I use len function. It's much faster than empty(). len(df.index) is even faster.

import pandas as pd
import numpy as np

df = pd.DataFrame(np.random.randn(10000, 4), columns=list('ABCD'))

def empty(df):
    return df.empty

def lenz(df):
    return len(df) == 0

def lenzi(df):
    return len(df.index) == 0

'''
%timeit empty(df)
%timeit lenz(df)
%timeit lenzi(df)

10000 loops, best of 3: 13.9 µs per loop
100000 loops, best of 3: 2.34 µs per loop
1000000 loops, best of 3: 695 ns per loop

len on index seems to be faster
'''

Answer 2

I use this to check if a particular column in a dataFrame has no values or is empty:

len(df.col_name.value_counts()) > 0

Answer 3

According to the Pandas documentation you need to use the empty property and not isempty

E.g.

In [12]: df.empty
Out[13]: False

Answer 4

If NDFrame contains only NaNs, it is still not considered empty. See the example below.

Examples

An example of an actual empty DataFrame. Notice the index is empty:

>>> df_empty = pd.DataFrame({'A' : []})
>>> df_empty
Empty DataFrame
Columns: [A]
Index: []
>>> df_empty.empty
True

If we only have NaNs in our DataFrame, it is not considered empty! We will need to drop the NaNs to make the DataFrame empty:

>>> df = pd.DataFrame({'A' : [np.nan]})
>>> df
    A
0 NaN
>>> df.empty
False
>>> df.dropna().empty
True

Source

Answer 5

Long answer: it depends

Depending on your definition of empty, your answer may vary a lot, as indicated by the various other answers. I try to summarize, but first have some test DataFrames:

no_rows = pd.DataFrame([], columns=list('ABCD'))
no_cols = pd.DataFrame([], index=range(3))
only_na = pd.DataFrame(float('nan'), index=range(3), columns=list('ABCD'))

Empty == no rows

The currently most popular answer takes this approach: a DataFrame with 0 rows is empty:

def empty_no_rows(df):
    return len(df.index) == 0

Empty == no columns

Not mentioned yet, but equally valid would be the transposed definition:

def empty_no_cols(df):
    return len(df.columns) == 0

Empty == no values

No actually, what you care about are values! If you prefer a definition that can deal with both empty index or columns, the following definition would work:

def empty_no_vals(df):
    return df.values.size == 0

Empty == what pandas says

Why not live with pandas' own definition of emptiness, which for these test cases leads to the same results as the no values definition:

def empty_native(df):
    return df.empty

Pandas' own implementation basically just checks if len(df.columns) == 0 or len(df.index) == 0, and never looks at values directly.

Empty == not only NaN's

Finally, you might want to ignore NaN in your considerations:

def empty_nans(df):
    return df.dropna(how='all').empty

But actually, this opens the next can of worms, as you now must decide how and along which axis you want to discard values? I stick to the more conservative all, here. And once those values are dropped, you could now apply all of the above definitions to its result.

Comparison

DataFrame	empty_no_rows	empty_no_cols	empty_no_vals	empty_native	empty_nans
no_rows	✅ True	❌ False	✅ True	✅ True	✅ True
no_cols	❌ False	✅ True	✅ True	✅ True	✅ True
only_na	❌ False	❌ False	❌ False	❌ False	✅ True

_{Editorial remark: I would call all those functions is_empty_..., but that leads to a comparison table that is too wide.}

Answer 6

Thanks @sparrow I used this to test for datetime columns:

    if len(df.select_dtypes(include='datetime').iloc[0].value_counts()) == 0:
        print('DF DATETIME COLUMNS: ', len(df_dt.iloc[0].value_counts()))

None of the other methodes (a.any(), a.empty()...) worked. select returns with a non-empty index but with empty columns so I think that's it. I think it actually returns a series, hence the zero iloc.

Answer 7

To check if a serie is empty I do the following:

df['your data'].notna().sum() == 0

Answer 8

I will explain my experiment:

I had a code as following:

matched_this = pd.Series(matched_groups)

And sometimes matched_this was []. But match_this.empty was False. So to solve this problem I used as follows:

if match_this[0]:
  # do ...

Answer 9

Q: How to check if pandas Series is empty?

Many answers here are dealing with measuring an empty pandas dataframe. A pandas dataframe is not the same as a pandas Series. An individual pandas Series may change its length during a data manipulation process. It may be useful to verify the length of a Series directly, using either

1. Series.empty
2. len()
3. len(Series.array)

Short examples

Let's create 3 dataframes to compare the output when measuring the length of a Series

>>> df0 = pd.DataFrame({'X' : []})
>>> df0
Empty DataFrame
Columns: [X]
Index: []

>>> df1 = pd.DataFrame({'A' : ['np.Nan']})
>>> df1
   A
0  np.NaN

>>> df2 = pd.DataFrame({'B' : ['b']})
>>> df2
   B
0  b

The pandas Series df0.X is empty. Therefore,

1. >>> df0.X.empty
   True
2. >>> len(df0.X)
   0
3. >>> len(df0.x.array)
   0

Each other pandas Series pf1.A and df2.B contains 1 value. Therefore,

df1.A                       |  df2.B
----------------------------|----------------------------
1. >>> df1.A.empty          |   1. >>> df2.B.empty
   False                    |      False
2. >>> len(df1.A)           |   2. >>> len(df2.B)
   1                        |      1
3. >>> len(df1.A.array)     |   3. >>> len(df2.B.array)
   1                        |      1

Hence, to verify if a pandas Series is empty, e.g. for df['A'], one may use

if df.A.empty == True:

if len(df.A) == 0:

if len(df.A.array) == 0:

Answer 10

If you want to check the columns id you can try df[df["id"].isna()].shape[0] == 0