CODEWITHSUNDEEP
pythonpandas
valenzio
valenzio

Reputation: 823

Python: Get postion in pandas dataframe

Hi folks I have this simple problem, where I am trying to find the iloc row postion of all NaN field in a dataframe (column B). So far I am solving the problem the following way:

rng= ['AA', 'BB', 2, 3, 4, 5]
df1 = pd.DataFrame(np.random.randn(6, 3), index=rng, columns=['A', 'B', 'C'])
df1.iloc[1][1]= np.nan

+----------------------------------+
|            A         B         C |
+----------------------------------+
| AA  0.198267 -1.469309 -1.751756 |
| BB -1.376388       Nan  0.988391 |
| 2  -1.697636 -0.814975  0.614170 |
| 3  -1.187977  1.240791 -1.079049 |
| 4  -1.495139  0.215619 -1.572205 |
| 5   1.157736 -0.656647 -0.307207 |
+----------------------------------+

ind_com=df1.loc[df1.B.isnull()].index.values.tolist()
ind_list=[]
for ii in ind_com:
    ind_list.append(df_temp.index.get_loc(ii))

ind_list = 1 Surely there must be a better way. Thank you

Upvotes: 1

Views: 65

Answers (3)

thomaskeefe
thomaskeefe

Reputation: 2394

Since the iloc really just a 0-indexed row number, it should be identical to enumerating over the nulls in the column:

[iloc for iloc, null in enumerate(df['B'].isnull()) if null]

Upvotes: 1

jezrael
jezrael

Reputation: 863236

I think you need:

pos = [df1.index.get_loc(x) for x in df1.index[df1.B.isnull()]]

Another solution with numpy.where:

pos = np.where(df1.B.isnull().values)[0].tolist()

Or numpy.nonzero:

pos = np.nonzero(df1.B.isnull().values)[0].tolist()

Upvotes: 2

Scott Boston
Scott Boston

Reputation: 153500

Let's a different approach (no looping):

ind_list = np.where(df1.B.isnull().as_matrix())[0].tolist()
print(ind_list)

Upvotes: 2

Related Questions