CODEWITHSUNDEEP
python-3.xpandas
user9431057
user9431057

Reputation: 1253

Get the index of NaN values from a selected set of rows

I have a data frame like this,

    ID   Cus_ID cl_id
0   5.0  200    0
1   NaN  200    0
2   NaN  200    1
3   14.0 200    2
4   15.0 200    2
5   16.0 200    2
6   NaN  200    3

From dataframe above, I am want to extract rows 0 through 4 and check if there is any values in 'ID' column has a NaN values. I tried this,

rows_needed = [0,1,2,3,4]

df.iloc[rows_needed,0].isnull().index.tolist()

But I get the following,

[0, 1, 2, 3, 4]

I am expecting to get the indexes of [1,2]. How can I get my desired output?

When I do this,

df.iloc[rows_needed,0].isnull()

I get,

0    False
1     True
2     True
3    False
4    False
Name: ID, dtype: bool

Not sure where I am making the mistake not to get my output.

Upvotes: 3

Views: 596

Answers (4)

Andy L.
Andy L.

Reputation: 25249

You may use index.symmetric_different with dropna to find those index is not NaN as follows:

df.iloc[rows_needed,0].dropna().index.symmetric_difference(rows_needed).tolist()

or 

(df.iloc[rows_needed,0].dropna().index ^ rows_needed).tolist()

Out[684]: [1, 2]

Upvotes: 1

user3483203
user3483203

Reputation: 51155

Two steps for clarity. Slice, then mask based on that slice.

u = df.iloc[rows_needed, 0]

u[u.isnull()].index.tolist()


[1, 2]

Upvotes: 4

Umar.H
Umar.H

Reputation: 23099

You're very very close, what you need to do is chain .iloc and .loc with a ==TRUE to get your result

your_indices = (df.iloc[rows_needed]
                .loc[df.ID.isnull()==True]
                .index.tolist())

print(your_indices)
[1, 2]

Upvotes: 6

BENY
BENY

Reputation: 323276

Let us chain it loc will only pick the result yield True out 

rows_needed = [0,1,2,3,4]
df.iloc[rows_needed,0].isnull().loc[lambda x : x].index.tolist()
Out[240]: [1, 2]

Upvotes: 3

Related Questions