Skip to content
BENIBIX Learn

Exploratory Data Analysis (EDA)

Data📊 Analytics & BIData Analysis

Also known as Exploratory Analysis

The first look at new data. You check its shape, its gaps and its odd values.

Think of it like

It is like walking through a new house. You open each door to see what is there.

Example

Emma gets a file of 980 customers. In ten minutes she finds the city is blank for 327 of them, and one age says 214.

Why it matters

Many wrong reports come from data nobody looked at. A short look first saves days of fixing later.

Definition

Exploratory data analysis is the first look at a new dataset. You count the rows and see how the values spread out. You check what is missing and what looks odd. Quick charts and simple sums do most of the work. The goal is to know the data before you trust any answer from it.

Example code

pycon
>>> df.isna().sum()
price      0
city     327
email      0
One line counts the blanks in each column. The city column has many.
Three first looks: how values spread, what is missing, and what does not fit.

Common mistake

Skipping it because the data came from a trusted team. Every new file needs a look of its own.

Key takeaways

  1. 1
    Count first. Rows, columns and the blanks in each.
  2. 2
    Plot it. A quick chart shows what a table hides.
  3. 3
    Ask what is odd. Find values that cannot be true.