Recently, I have started to realize that I have a growing interest in DSA and similar topics.
Although I am unsure of where this stems from, I have started to also increasingly take a liking to
competitive programming. Furthermore I have begun my journey on Code Forces! I do have some experience with
online judges such as Kattis or LeetCode, but Code Forces seems like a great place to train my problem
solving skills further. For ICPC or other formal competition prep I may choose to stick with Kattis though.
The goal for me on Code Forces is to reach CM (Candidate Master), which sits around ~1900 elo points
and slots you a spot in the top 5000 users on the website. I understand this is quite ambitions and at
the moment I am doing terrible! This is okay though. I currently have a modest 778 elo points
and the title newbie, indicating that I suck at competitive programming (ranked 107488 globally).
However, the goal at the moment is not the elo rating, nor should it be. If I really want to reach my goal of CM,
I will first need to study quite a lot of DSA, and I look forward to the journey ahead of me.
There is no plan!
If I take this too seriously then I will most likely fall out of love with the process. I am dedicating quite a bit of time
to understanding the underlying concepts behind problems though. I've gotten my hands on a copy of the book CP4,
which is a text written with ICPC candidates in mind. It walks you through the basics of competitive programming, such as what language to use for certain problems, all the way
to rare or uncommon algorithms that you may not even see in your competitive programming career. One of my favorite parts of this text thus far is the
problems they supply you with for each topic. At the end of a section, the authors have kindly left problems that are in real repositories such as
Kattis or UVa (Online Judge). I also plan to continue doing the
CSES problem set for some fundamentals, as I am a firm believer in curated lists of problems for times when you aren't sure
how to proceed with your studies.
Let's quantify my current level a bit more. As previously mentioned, I am rated 778 on Code Forces, having competed in 3 contests since starting with the website. A contest usually ends for me
after solving the first problem in an unholy amount of time, and then understanding the second problem, but failing to code it due to missing a condition or two during the writing process.
This is actually very apparent in my rankings, with my lowest ranking in a contest being in a Div. 3 competition, in which I placed 22502. Again though, I think this is great. Once the initial
elo boosting that comes from merely participating wears off, I will have nothing but my own improvement to bank on, or else I will remain in the newbie bracket. If I continue to improve, my elo
will rise naturally. I can generally solve 800 - 1000 rated problems in a low pressure setting, but will occasionally look at editorials to give me some guidance if I am stuck for too long.
On Kattis, I have found the style that problems are written in is much more approachable than Code Forces. This may be a bit more apparent in my rating, as my current rank is 13472 globally, and
I have a score of 107.6 (prev. 110.1). The breadth of problems that I have solved on Kattis is also a bit wider as well, having experimented with basic DP, some bit manipulation, and a linear
recurrence problem that required matrix multiplication for fast calculation (my first Kattis hard!). There are no contests on Kattis though, or at least not on the scale of Code Forces as most are
invitational. However, when I return back to Canada later this year, we will mainly use Kattis for both training and competing in competitive programming contests.
LeetCode I haven't touched in a while. With 61 solved problems and a contest rating of 1476 (globally 366906/711693), it is evident I'm not doing too hot. I don't find leetcode to be very interesting,
but I might come back to it after a while to see if I have improved at all!
As for some other sites that people may know (or that I have previously mentioned), I have solved 6/400 on CSES (very low, but up from here I promise), and on AtCoder I currently have no rating at all,
but I plan to participate in a contest to change this soon.
Now we will take a look at a problem I have solved from a Code Forces contest, and keep in mind I struggled to solve this lol.
Problem: 2121A, Letter Home
You are given an array of distinct integers \( x_1, x_2, ... x_n\) and an integer \(s\).
Initially, you are at position \(pos = s\) on the \(X\) axis. In one step, you can perform exactly one of the following two actions:
This problem is really easy. But guess how long it took me to solve it! Almost 2 hours. 2 hours!!! To make things worse, my code is a little bit ugly.
#include <bits/stdc++.h>
using namespace std;
void solve() {
int n,s,x; // s = start, n = len
int l = 101, r = -1;
cin>>n>>s;
for (int i=0;i>n;i++) {
cin>>x;
l = min(x, l); r = max(x, r);
}
if (s >= r) {
cout<<s-l<<'\n';
} else if (s <= l) {
cout<<r-s<<'\n';
} else if (s - l <= r - s) {
cout<<s - l + r - l<<'\n';
} else {
cout<<r - s + r - l<<'\n';
}
int main() {
int t;
cin>>t;
while (t--) solve();
return 0;
}
This is my first real post on this blog, and I would like to end it here! If you have read this far, I extend you my thanks. If all goes well, maybe
one day you will come back here to see me realize this ambitious goal I have set for myself.
Until next time!
- Liam