Leetcode 547 solving with intuition - 6/8/26

bala — Mon, 08 Jun 2026 15:17:33 +0000

Q1) 547. Number of Provinces

💡 Initial Idea & The Mistake

My first instinct was to simply count the rows where the array sum equals 1 (meaning a city is only connected to itself), assuming everything else would naturally hook together into a single province.

The flaw in that logic: I completely overlooked scenarios where you might have multiple distinct, disconnected clusters of cities (e.g., two cities connected to each other, while three other cities form their own separate group).

To fix this, I broke out a notepad, sketched out the graph visually, and realized this was a classic graph traversal problem perfectly suited for Depth-First Search (DFS).

🛠️ Optimized DFS Solution (Java)

In my initial draft, I used an external helper function (allVisit) to scan the visited array repeatedly inside a loop. This added unnecessary overhead. By simply looping through all cities once from 0 to n-1 and triggering a DFS whenever we hit an unvisited city, we can count the provinces much more efficiently.

class Solution {
    public int findCircleNum(int[][] isConnected) {
        int cities = isConnected.length;
        int provinces = 0;
        boolean[] visited = new boolean[cities];

        // Loop through each city. If it hasn't been visited, 
        // it represents the start of a new province.
        for (int i = 0; i < cities; i++) {
            if (!visited[i]) {
                dfs(i, visited, isConnected);
                provinces++; // Increment the province count
            }
        }

        return provinces;
    }

    private void dfs(int city, boolean[] visited, int[][] matrix) {
        visited[city] = true;

        // Visit all adjacent, unvisited cities
        for (int i = 0; i < visited.length; i++) {
            if (matrix[city][i] == 1 && !visited[i]) {
                dfs(i, visited, matrix);
            }
        }
    }
}

I'm Starting Over — And That Might Be My Biggest Advantage

bala — Sat, 06 Jun 2026 13:47:08 +0000

Yesterday, I finished my 6th semester examinations.

For most students, that's the end of a stressful academic phase. For me, it's the start of something more demanding — placement season.

Here's a realization I've been sitting with:

Most people think placement success comes from knowing more. I think it comes from being willing to start over.

The Comfortable Trap

I've spent the last few years building projects, solving problems, exploring AI, and competing in hackathons. My DSA skills are solid — not exceptional, but not weak either.

The dangerous thing about being "decent" is that it's easy to stop questioning yourself. You rely on old knowledge, old approaches, old confidence.

Placement prep forced me to confront an uncomfortable truth: what got me here may not be enough to get me where I want to go.

So instead of defending what I already know, I'm choosing to play the game again — this time with experience.

The Context Nobody Talks About

I'm from Electrical and Electronics Engineering, not Computer Science. I don't have a curriculum built around software engineering, OS, databases, or networking.

That's fine.

Being outside the traditional path taught me something: when you don't have the roadmap, you learn to build one. Every project, every debugging session, every late night has shown me that progress is rarely about ideal circumstances — it's about consistency despite imperfect ones.

The Plan

For the next few months, I'm committing to six areas:

DSA — Pattern recognition and problem-solving under pressure, not just volume
Projects — Building things that strengthen engineering thinking, not just resume lines
Aptitude — Consistent practice over last-minute cramming
Communication — Explaining ideas clearly is as critical as understanding them
Building in public — Accountability creates momentum; feedback creates improvement
Networking — Opportunities find people who position themselves well

The Real Goal

Not perfection across any of these. Just Kaizen — 1% better every day.

The person who walks into interviews a few months from now won't be the same person who finished his exams yesterday. That's the point.

DEV Community: bala