CODEWITHSUNDEEP
pythonpython-3.xpandasdataframe
Dave
Dave

Reputation: 2591

How to get value from a row in Pandas DataFrame?

I have a pandas data frame that contains one row. I know what the column names are in this row. I would like to pull the value (just the value, not the type, or other metadata) from each cell in this row. How do I do this? I am using Python 3.

I have tried the following, but it always fails, because you can't use a column.

print(data_frame.head().values['ColumnName'])

So then I tried this, which gives me the value of ColumnName, but then crashes.

print(data_frame.iloc[0]['ColumnName'])

File "/usr/local/lib/python3.8/dist-packages/pandas/core/indexing.py", line 895, in getitem return self._getitem_axis(maybe_callable, axis=axis) File "/usr/local/lib/python3.8/dist-packages/pandas/core/indexing.py", line 1501, in _getitem_axis self._validate_integer(key, axis) File "/usr/local/lib/python3.8/dist-packages/pandas/core/indexing.py", line 1444, in _validate_integer raise IndexError("single positional indexer is out-of-bounds") IndexError: single positional indexer is out-of-bounds

I can get it to work if I loop through it, but I don't need or want this since I only have one row.

for x in data_frame.itertuples(index=False):
     print(x.ColumnName1)
     print(x.ColumnName2)

I also was able to get something to give me the value along with a bunch of other metadata, but unfortunately I can't recall what code I used to get that.

All I want to do is something like this and get the value of ColumnName.

data_frame[0]['ColumnName']

Is this possible?

Edit Here is the code. It calls a stored procedure in a SQL database.

    query = "EXEC p_get_data @id = '{0}'".format(id)
connection_string = get_connection_string()
engine = sqlalchemy.create_engine(connection_string)
data_frame = pd.read_sql_query(query, engine)
data_frame.reset_index(inplace=True)

#print(data_frame.loc[0]['ColumnName1'])
#print(data_frame.loc[0, 'ColumnName1'])
#print(data_frame['ColumnName1'].values[0])
#print(data_frame['ColumnName1'].iloc[0])
#print(data_frame.iat[0, data_frame.columns.get_loc('ColumnName1')])
print(data_frame['ColumnName1'][0] )

The stored procedure

ALTER PROCEDURE [dbo].[p_get_data]
    @id varchar(200)
AS
    SELECT A.ColumnName1, E.ColumnName2 
    FROM E (NOLOCK)
    INNER JOIN A (NOLOCK) ON E.PrimaryKey = A.PrimaryKey
    WHERE Id = @id
      AND A.Type = 'test'

The data looks like this: (ColumnName1 is an integer, ColumnName2 is a varchar

ColumnName1     ColumnName2
---------------------------
1               ABCDE1234

Upvotes: 0

Views: 4385

Answers (2)

Cedric Zoppolo
Cedric Zoppolo

Reputation: 4743

Toy example

Lets say you have following DataFrame:

In [1]: import pandas as pd
In [5]: df = pd.DataFrame({"ColumnName1":[1],"ColumnName2":['text']})

Then you have:

In [6]: df
Out[6]: 
   ColumnName1 ColumnName2
0            1        text

Values from single row

If you want to get the values from first row you just need to use:

In [9]: df.iloc[0]
Out[9]: 
ColumnName1       1
ColumnName2    text
Name: 0, dtype: object

Or:

In [10]: df.iloc[0,:]
Out[10]: 
ColumnName1       1
ColumnName2    text
Name: 0, dtype: object

And if you want to get an array instead you can use:

In [11]: df.iloc[0].values
Out[11]: array([1, 'text'], dtype=object)

Values from single cell

If you want the values from a single cell just use:

In [14]: df.iloc[0,0]
Out[14]: 1

In [15]: df.iloc[0,1]
Out[15]: 'text'

Or even with:

In [22]: df.ColumnName1[0]
Out[22]: 1

In [23]: df.ColumnName2[0]
Out[23]: 'text'

Upvotes: 0

ThePyGuy
ThePyGuy

Reputation: 18406

If it's single row, You should be fine using something like this:

>>> data_frame['colName'].values[0]
Out[8]: 1

Or,

>>> data_frame['colName'].iloc[0]
Out[9]: 1

Or,

>>> data_frame.iloc[0, data_frame.columns.get_loc('colName')]
Out[11]: 1

Or,

>>> data_frame.iat[0, data_frame.columns.get_loc('colName')]
Out[12]: 1

Or,

>>> data_frame.loc[0, 'colName']
Out[13]: 1

Upvotes: 1

Related Questions