Lesson 07

DataFrame Attributes

Read the shape of a table with shape, size, index, columns, axes and dtypes, then look inside with head, tail, info, and describe.

Intermediate25 min

DataFrame Attributes hero Module 24 Lesson 7

Before you analyse a table, find out what it is. How many rows and columns, what the columns are called, what type each one holds, and whether anything is missing. These five calls are the first thing to run on new data.

The five calls that describe a new table

Size and labels

shape is (rows, columns), so shape[0] is the row count. size is every cell. index and columns hold the two sets of labels, and axes is both of them in a list.

python

import pandas as pd

learners = pd.DataFrame({
    "name": ["Avery", "Jordan", "Casey", "Riley", "Morgan", "Quinn"],
    "track": ["Python", "NumPy", "Python", "pandas", "NumPy", "pandas"],
    "city": ["Northgate", "Eastport", "Northgate", "Westfield", "Eastport", "Westfield"],
    "hours": [42, 31, 55, 24, 47, 38],
    "score": [88, 74, 91, 65, 82, 79],
})
print(learners.shape, learners.size, learners.ndim)
print("rows:", learners.shape[0], "columns:", learners.shape[1])
print(learners.columns.tolist())
print(learners.index)

Output:

text

(6, 5) 30 2
rows: 6 columns: 5
['name', 'track', 'city', 'hours', 'score']
RangeIndex(start=0, stop=6, step=1)

Column types

dtypes gives one type per column. Text columns show as object. A numeric column that holds a missing value becomes float64, because NaN is a float.

python

import pandas as pd

learners = pd.DataFrame({
    "name": ["Avery", "Jordan", "Casey", "Riley", "Morgan", "Quinn"],
    "track": ["Python", "NumPy", "Python", "pandas", "NumPy", "pandas"],
    "city": ["Northgate", "Eastport", "Northgate", "Westfield", "Eastport", "Westfield"],
    "hours": [42, 31, 55, 24, 47, 38],
    "score": [88, 74, 91, 65, 82, 79],
})
print(learners.dtypes)
print(learners.select_dtypes("number").columns.tolist())

Output:

text

name     object
track    object
city     object
hours     int64
score     int64
dtype: object
['hours', 'score']

info

info() prints the row count, each column with its non-null count and type, and the memory used. The non-null counts are the quickest way to spot missing data.

python

import pandas as pd

learners = pd.DataFrame({
    "name": ["Avery", "Jordan", "Casey", "Riley", "Morgan", "Quinn"],
    "track": ["Python", "NumPy", "Python", "pandas", "NumPy", "pandas"],
    "city": ["Northgate", "Eastport", "Northgate", "Westfield", "Eastport", "Westfield"],
    "hours": [42, 31, 55, 24, 47, 38],
    "score": [88, 74, 91, 65, 82, 79],
})
learners.info()

Output:

text

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 6 entries, 0 to 5
Data columns (total 5 columns):
 #   Column  Non-Null Count  Dtype 
---  ------  --------------  ----- 
 0   name    6 non-null      object
 1   track   6 non-null      object
 2   city    6 non-null      object
 3   hours   6 non-null      int64 
 4   score   6 non-null      int64 
dtypes: int64(2), object(3)
memory usage: 372.0+ bytes

describe

describe() summarizes the numeric columns. Pass include="all" to add the text columns, which report their distinct count and most common value instead of a mean.

python

import pandas as pd

learners = pd.DataFrame({
    "name": ["Avery", "Jordan", "Casey", "Riley", "Morgan", "Quinn"],
    "track": ["Python", "NumPy", "Python", "pandas", "NumPy", "pandas"],
    "city": ["Northgate", "Eastport", "Northgate", "Westfield", "Eastport", "Westfield"],
    "hours": [42, 31, 55, 24, 47, 38],
    "score": [88, 74, 91, 65, 82, 79],
})
print(learners.describe())
print(learners.describe(include="object"))

Output:

text

           hours      score
count   6.000000   6.000000
mean   39.500000  79.833333
std    11.113055   9.495613
min    24.000000  65.000000
25%    32.750000  75.250000
50%    40.000000  80.500000
75%    45.750000  86.500000
max    55.000000  91.000000
         name   track       city
count       6       6          6
unique      6       3          3
top     Avery  Python  Northgate
freq        1       2          2

First and last rows

head() and tail() show five rows by default. On a real file of thousands of rows, this is how you look before you leap.

python

import pandas as pd

learners = pd.DataFrame({
    "name": ["Avery", "Jordan", "Casey", "Riley", "Morgan", "Quinn"],
    "track": ["Python", "NumPy", "Python", "pandas", "NumPy", "pandas"],
    "city": ["Northgate", "Eastport", "Northgate", "Westfield", "Eastport", "Westfield"],
    "hours": [42, 31, 55, 24, 47, 38],
    "score": [88, 74, 91, 65, 82, 79],
})
print(learners.head(2))
print(learners.tail(2))

Output:

text

     name   track       city  hours  score
0   Avery  Python  Northgate     42     88
1  Jordan   NumPy   Eastport     31     74
     name   track       city  hours  score
4  Morgan   NumPy   Eastport     47     82
5   Quinn  pandas  Westfield     38     79

Key takeaways

shape is (rows, columns) and size is every cell
Text columns show as object in dtypes
info() shows non-null counts, so missing data stands out
describe() covers numbers unless you include the rest
Next: Selecting Columns and Rows
Previous: Creating a DataFrame

Loading playground…

Check your understanding

DataFrame Attributes Quiz

What does df.shape return?

1 / 10

← Creating a DataFrame

Selecting Columns and Rows →