Summaries and scaling सारांश और स्केलिंग

1. Mean, min and max

Before training anything, look at your data. Three quick summaries of a column:

• mean (average): add everything, divide by how many
• min: the smallest value
• max: the largest value

They catch mistakes early. If the max weight of a mango is 9000 g, someone typed an extra zero!

To find the max, start with the first value and walk through the rest, keeping the biggest you have seen.

                const marks = [62, 75, 48, 90];
let total = 0;
for (const m of marks) total += m;
console.log("mean", total / marks.length);
              

Challenge: Report-card summary

Write three functions that take an array of numbers:

• mean(xs) returns the average
• minOf(xs) returns the smallest value
• maxOf(xs) returns the largest value

You may use a loop or Math.min(...xs). Arrays will never be empty.

2. Normalising: squeeze into 0–1

Weight goes from 50 to 300 grams, sweetness from 1 to 10. If we compare fruits with these raw numbers, weight shouts and sweetness whispers — a 10-gram difference looks bigger than the whole sweetness scale!

Normalising puts every feature on the same 0–1 scale:

scaled = (x - min) / (max - min)

• the smallest value becomes 0
• the largest becomes 1
• everything else lands in between

Careful: if all values are equal, max - min is 0 and you would divide by zero. In that case return 0 for every value.

                const w = [50, 175, 300];
const lo = Math.min(...w), hi = Math.max(...w);
console.log(w.map((x) => (x - lo) / (hi - lo)));
              

Challenge: Scale it to 0–1

Write normalise(xs) that returns a new array where each value is (x - min) / (max - min). If all values are the same, return an array of zeros (same length).