Hello everyone, and welcome to my first blog!

I’m very excited to begin this journey. Since childhood, I’ve been deeply fascinated by astronomy, and today, I finally get to express that passion through writing and sharing my learning experiences with such a great community.

A Memorable Start – Meeting My Mentors

This week started off on a truly inspiring note. On May 23rd, I had the opportunity to meet my mentors—individuals with deep expertise and experience in the field. At first, I was a bit nervous and hesitant. But the moment the session began, all my fear started fading away.

We began with short introductions and then quickly moved on to discussing plans for the upcoming work. I had a few doubts—some of which I felt were too basic or weird to ask—but I’m so glad I did! My mentors answered them with such clarity and patience. They even gave me valuable suggestions for improving one of my current functions (called recpiec) and helped us finalize a workflow to proceed with.

Learning Through Articles and Real Research

Later in the week, I spent time diving deep into articles to strengthen my understanding. One article that particularly stood out was authored by one of my mentors:

“No Time for Dead Time: Use the Fourier Amplitude Differences to Normalize Dead-time-affected Periodograms”

This article taught me about the problem of dead time in time-series data, especially when analyzing periodic signals using periodograms. Dead time refers to gaps or inactive intervals in data collection, which can significantly distort frequency analysis. The article introduced a technique using Fourier Amplitude Differences to correctly normalize these periodograms, ensuring the results remain accurate and meaningful. This method was new to me and absolutely fascinating!

Exploring Bartlett’s Method and Hanning Windows

Along with that, I also explored Bartlett's method—a signal processing technique used to reduce the variance of power spectral density estimates. It involves dividing the signal into non-overlapping segments, computing a periodogram for each, and averaging them. This approach provides a smoother, more reliable estimate of the spectrum, which is especially useful in noisy datasets (which are quite common in astronomy!).

Inspired by this, I also implemented the Hann window in my code to taper the signal before applying spectral analysis. This windowing technique helps minimize spectral leakage by reducing the discontinuities at the edges of the signal. I was surprised to see how much of a difference it made in the clarity of results!

Here’s a quick snippet showing how I used the Hann window:

function hanning(N::Int)
		return [0.5 * (1 - cos(2π * n / (N-1))) for n in 0:N-1]
		end

I have pasted the plots also to get more insights :)

Oh! u have read my entire blog, thank u :)