Create a single pandas.core.frame.Pandas object

user_1179317 0
Viewed 520

How do I create a single row of data as a pandas.core.frame.Pandas object?

So, when you iterate a dataframe (df) with for row in df.itertuples(), each row will be a pandas.core.frame.Pandas object. I want to create an object of that type.

For example:

import pandas as pd

d = [{'a': 1, 'b': 2}]
df = pd.DataFrame(d)
print(type(df))
print("a", df.a)

for row in df.itertuples():
    print(type(row))
    print("a", row.a)


myrow = None
df = pd.DataFrame(d)
for row in df.itertuples():
    myrow = row
print(type(myrow))
print("a", myrow.a)

Its output is:

<class 'pandas.core.frame.DataFrame'>
a 0    1
Name: a, dtype: int64
<class 'pandas.core.frame.Pandas'>
a 1
<class 'pandas.core.frame.Pandas'>
a 1

As you can see, pandas.core.frame.DataFrame acts differenctly as pandas.core.frame.Pandas, as expected.

I dont really want to create myrow the way I did above, so I was wondering whats the more efficient/direct way to create that object. Assuming I only have 1 list of data I want to convert to pandas.core.frame.Pandas object

pandas dataframe python
2 Answers

Explanation:

Every iteration of itertuples gives an object like:

Pandas(Index=0, a=1, b=2)

Pandas is just the default name of the itertuples name argument, example:

>>> help(df.itertuples)
Help on method itertuples in module pandas.core.frame:

itertuples(index=True, name='Pandas') method of pandas.core.frame.DataFrame instance
    Iterate over DataFrame rows as namedtuples.
    
    Parameters
    ----------
    index : bool, default True
        If True, return the index as the first element of the tuple.
    name : str or None, default "Pandas"
        The name of the returned namedtuples or None to return regular
        tuples.

The default name is Pandas. That is just the name of the namedtuple.

If you change that name, the types would become different:

for row in df.itertuples(name='newname'):
    print(row)
    print(type(row))
    print("a", row.a)

Output:

newname(Index=0, a=1, b=2)
<class 'pandas.core.frame.newname'>
a 1

Replication:

This is the default behavior of namedtuples, example:

>>> from collections import namedtuple
>>> a = namedtuple('Pandas', ['x', 'y', 'z'])
>>> a(1, 2, 3)
Pandas(x=1, y=2, z=3)
>>> type(_)
<class '__main__.Pandas'>
>>>

As you can see, the type of this is Pandas, which is the type name of this namedtuple, so itertuples doesn't give a Pandas object, instead it's just the default set name of the itertuples namedtuple output.

Documentation reference:

As mentioned in the collections.namedtuple documentation:

Returns a new tuple subclass named typename. The new subclass is used to create tuple-like objects that have fields accessible by attribute lookup as well as being indexable and iterable. Instances of the subclass also have a helpful docstring (with typename and field_names) and a helpful __repr__() method which lists the tuple contents in a name=value format.

As you can see, it creates a new type.

user_87083640

To understand the thing better, I checked the source code. And it seems it's just the default name passed to the collections.namedtuple() factory to create a named tuple. This line also emphasizes what I said.

Here is a funny example to prove what I said:

import pandas as pd

d = [{'a': 1, 'b': 2}]
df = pd.DataFrame(d)
print(type(df))
print("a", df.a)

for row in df.itertuples():
    print(type(row))
    print("a", row.a)


myrow = None
df = pd.DataFrame(d)
for row in df.itertuples(name="user"):
    myrow = row
print(type(myrow))
print("a", myrow.a)

Output

<class 'pandas.core.frame.DataFrame'>
a 0    1
Name: a, dtype: int64
<class 'pandas.core.frame.Pandas'>
a 1
<class 'pandas.core.frame.user'>
a 1
user_37961130
Related