Reading Data

In real work you load data far more often than you type it in by hand. pandas can read CSV, Excel, JSON, and SQL — this lesson focuses on read_csv, the one you'll reach for constantly.

Reading a CSV file

Given a file sales.csv sitting next to your script:

order_id,product,quantity,price
1,Notebook,4,3.50
2,Pen,10,1.20
3,Notebook,2,3.50

import pandas as pd

df = pd.read_csv("sales.csv")
print(df)

   order_id   product  quantity  price
0         1  Notebook         4    3.5
1         2       Pen        10    1.2
2         3  Notebook         2    3.5

The first line of the file became the column names automatically, and pandas guessed a sensible dtype for each column — whole numbers for order_id and quantity, decimals for price.

Inspecting a loaded DataFrame

Two methods worth running on reflex right after loading anything real: .info() for a structural summary, and .describe() for quick statistics on the numeric columns:

print(df.info())
print(df.describe())

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 3 entries, 0 to 2
Data columns (total 4 columns):
 #   Column    Non-Null Count  Dtype
---  ------    --------------  -----
 0   order_id  3 non-null      int64
 1   product   3 non-null      object
 2   quantity  3 non-null      int64
 3   price     3 non-null      float64

       order_id  quantity  price
count       3.0       3.0    3.0
mean        2.0       5.3    2.7
std         1.0       4.2    1.3
...

.info() tells you the row count, each column's dtype, and how many non-missing values it has — the fastest way to spot missing data before it causes a problem three steps later. .describe() gives count, mean, standard deviation, min/max, and quartiles for every numeric column at once.