Lara Zijl 
Welcome to Day 2: Data Odyssey, our 365-day journey to master data science and AI, launched on Shivaratri, February 26, 2025! Yesterday, in Day 1: Data Odyssey – What is Data Science?, we defined data science as turning data into insights using statistics, programming, and context. We met Priya, our Delhi café owner, who used sales data to tweak her hours and menu. We outlined the workflow—collect, clean, analyze, act—and its real-world impact, setting our base. Today, we zoom into the core: What is data, and why is it the raw material of our craft?
Defining Data
Data is the heartbeat of data science—without it, there’s nothing to analyze. At its root, data is any piece of information we can capture, store, or process. It’s the world’s footprints, preserved in forms we can study: numbers (sales totals), text (customer reviews), images (weather maps), even audio (street sounds). Data isn’t new—ancient scribes logged harvests on clay—but today’s scale is colossal. Your phone alone generates gigabytes daily, from texts to GPS pings.
In Day 2: Data Odyssey, data is both treasure and challenge. It’s treasure because it holds answers: What’s popular? Who’s at risk? It’s a challenge because it’s raw—unorganized, incomplete, or noisy. Our task? Refine this rough ore into gold.
Types of Data
Data comes in flavors, and knowing them is key. Here’s the rundown:
-
Numerical Data – Numbers, the backbone of analysis.
-
Continuous – Any value (e.g., temperature: 23.7°C, 23.8°C).
-
Discrete – Whole numbers (e.g., cups sold: 5, 6).
-
Example: Priya’s revenue (₹500.75) or customers (12).
-
-
Categorical Data – Labels, not numbers.
-
Nominal – No order (e.g., items: chai, latte).
-
Ordinal – Ordered (e.g., ratings: low, medium, high).
-
Example: Priya’s menu or satisfaction scores.
-
-
Text Data – Words or phrases (e.g., “Loved the samosas!”).
-
Image Data – Pixels (e.g., café photos).
-
Time Series – Data over time (e.g., hourly sales: 8 AM, 9 AM).
Each type needs its own approach. Numbers crunch easily; text requires parsing; images demand special tools. Day 2: Data Odyssey kicks off this understanding.
Priya’s Data Up Close
Back to Priya’s café. Her data includes:
-
Numerical: Daily sales (₹12,345.50), cups sold (47).
-
Categorical: Items (chai, samosa), days (Monday, Tuesday).
-
Text: Notes (“New latte’s a hit!”).
-
Time Series: Sales by hour (8 AM: ₹500, 9 AM: ₹700).
She asks: What’s my peak hour? Numerical data (sales) and time series (hours) answer it. Later, text data (feedback) might explain why. Day 2: Data Odyssey will teach you to weave these together.
Where Data Originates
Data flows from endless wells:
-
Manual Entry – Priya typing sales into a spreadsheet.
-
Sensors – Thermometers, GPS, fitness trackers.
-
Digital Activity – Websites, apps, social media.
-
Public Records – Government stats, weather logs.
India’s data landscape is vast—railway bookings, monsoon records, census data. We’ll tap these as we progress.
Why Quality Counts
Data’s only as good as its accuracy. Picture Priya’s log with a typo: “₹5000” instead of “₹500.” Her busiest hour looks ten times busier! Bad data leads to bad calls—like opening at 4 AM for a ghost rush. Common pitfalls:
-
Missing Data – No sales for Wednesday.
-
Errors – “Chai” as “Cha.”
-
Outliers – A ₹50,000 sale (a glitch?).
Cleaning these is data science’s unsung hero. We’ll dive in soon.
Structured vs. Unstructured
Data has form:
-
Structured – Neat, like a table (Priya’s Date, Item, Price).
-
Unstructured – Wild, like emails, photos, videos.
Most data (80%+) is unstructured—think tweets or vlogs. Structured data’s simpler; unstructured needs extra finesse. Day 2: Data Odyssey prepares you for both.
Real-World Stakes
India’s weather agencies use monsoon data—numerical (rainfall), time series (daily totals), images (satellite)—to predict floods. A single error (a misread gauge) could miss a warning, risking lives. Quality data saves the day.
Recap Summary
Yesterday, Day 1: Data Odyssey defined data science as insight from data via stats, coding, and context—a cycle of collect, clean, analyze, act. Today, Day 2: Data Odyssey explored data: the raw material in types (numerical, categorical), sources (sensors, records), and the vital role of quality. It’s our fuel.
What’s Next
Tomorrow, in Day 3: Data Odyssey – How Do We Collect Data?, we’ll dive into data collection: How do we gather it? What tools or methods work? We’ll see how Priya could track sales smarter—and pitfalls to dodge. Join me there!
