<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Anilsarkar Phys on Medium]]></title>
        <description><![CDATA[Stories by Anilsarkar Phys on Medium]]></description>
        <link>https://medium.com/@anilsarkar.phys?source=rss-c9831424915d------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*3IL5ln-7hWUcqj4b</url>
            <title>Stories by Anilsarkar Phys on Medium</title>
            <link>https://medium.com/@anilsarkar.phys?source=rss-c9831424915d------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 02:08:16 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@anilsarkar.phys/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Beyond the Blackboard: Why the Future of Physics is Written in Code]]></title>
            <link>https://medium.com/@anilsarkar.phys/beyond-the-blackboard-why-the-future-of-physics-is-written-in-code-b580db44e080?source=rss-c9831424915d------2</link>
            <guid isPermaLink="false">https://medium.com/p/b580db44e080</guid>
            <category><![CDATA[science]]></category>
            <category><![CDATA[mathematics]]></category>
            <category><![CDATA[coding]]></category>
            <category><![CDATA[computational-physics]]></category>
            <category><![CDATA[physics]]></category>
            <dc:creator><![CDATA[Anilsarkar Phys]]></dc:creator>
            <pubDate>Sat, 06 Jun 2026 13:31:00 GMT</pubDate>
            <atom:updated>2026-06-06T13:31:00.746Z</atom:updated>
            <content:encoded><![CDATA[<h4>An intuitive introduction to numerical methods, approximation, and how computers solve the universe’s un-solvable equations.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/300/0*udVQ5pWQ_jbPwbje.jpg" /><figcaption>Molecular dynamics simulation of the oxidation of an aluminum nanoparticle.</figcaption></figure><p>Pop culture loves a specific image of a physicist: standing in front of a massive blackboard, chalk in hand, scribbling down a beautifully elegant equation that unlocks the secrets of the universe. It’s a romantic notion. But here is a dirty little secret of modern science: most of those beautiful equations cannot actually be solved by hand.</p><p>When you move past idealized textbook problems — like a perfect pendulum swinging in a vacuum — and enter the real world of chaotic weather systems, turbulent fluids, or colliding galaxies, exact mathematical solutions (what physicists call <em>analytical solutions</em>) completely vanish.</p><p>That is where numerical methods step in. They are the bridge between the elegant theories of physics and the gritty reality of computation.</p><h3>The Problem with the Real World</h3><p>In introductory physics, we learn Newton’s second law, <em>F = ma</em>, and use basic calculus to find exactly where a thrown baseball will land. But add air resistance that changes dynamically with speed, a shifting wind profile, and the rotation of the Earth, and the math quickly becomes an insurmountable wall.</p><p>Instead of looking for a perfect, all-encompassing algebraic formula, numerical methods ask a simpler, more pragmatic question: <em>Can we approximate the answer closely enough that the difference doesn’t matter?</em> By breaking down continuous, complex physics into millions of tiny, manageable mathematical steps, computers can simulate realities we could never calculate on paper.</p><h3>The Core Philosophy: Discretization</h3><p>To understand numerical methods, you must understand <strong>discretization</strong>. Nature is continuous; time flows smoothly, and space has no inherent grid lines. Computers, however, are inherently discrete. They handle distinct numbers, finite memory, and fixed steps.</p><p>Discretization is the process of carving up smooth, continuous physical phenomena into discrete pieces. Instead of trying to calculate the position of a planet at every infinitely small instant of time, a numerical simulation calculates its position at t = 0, then t = 0.01 seconds, then t = 0.02 seconds, and so on.</p><p>If the time step (often denoted as Δ t) is small enough, the discrete steps perfectly mimic continuous reality, keeping errors to an absolute minimum.</p><h3>The Toolkit: Three Classic Numerical Algorithms</h3><p>If you are just entering the world of computational physics, you will quickly cross paths with these three fundamental approaches:</p><ul><li><strong>Euler’s Method:</strong> The absolute starting point of numerical integration. Named after Leonhard Euler, this method uses the current state of a system and its rate of change to predict the next state. While incredibly intuitive, it suffers from accumulating errors over time, making it less ideal for long, precise simulations.</li><li><strong>Runge-Kutta Methods (Specifically RK4):</strong> Think of RK4 as Euler’s method on steroids. Instead of blindly marching forward based on a single trajectory data point, the RK4 algorithm takes four different sample steps across the time interval to calculate a highly accurate average trajectory. It is the undisputed workhorse of classical mechanics simulations.</li><li><strong>Monte Carlo Simulations:</strong> Named after the famous casino destination, this method relies on the power of randomness. If a physics problem has too many variables to calculate deterministically — like thousands of subatomic particles colliding in a detector — physicists use random sampling to find probabilistic outcomes.</li></ul><h3>Why This Matters Beyond the Lab</h3><p>Numerical physics isn’t just an academic exercise; it powers the modern infrastructure we rely on every day.</p><ul><li><strong>Meteorology:</strong> Global weather models solve fluid dynamics equations across a massive 3D grid of Earth’s atmosphere to predict where a hurricane will make landfall.</li><li><strong>Aerospace Engineering:</strong> Rocket trajectories and airflow over aircraft wings are simulated using computational fluid dynamics (CFD) long before a physical prototype is ever built.</li><li><strong>Astrophysics:</strong> Since we cannot collide two black holes in a physical laboratory, we simulate them using numerical relativity to predict the gravitational waves our detectors look for in deep space.</li></ul><h3>Final Thoughts: The Modern Physicist’s Toolkit</h3><p>The transition from analytic physics to numerical physics represents a profound shift in how we interrogate nature. It requires us to trade the aesthetic perfection of a closed-form equation for the pragmatic utility of an algorithmic approximation.</p><p>For the modern physicist, code is just as vital as calculus. By learning how to translate the laws of nature into loops, arrays, and algorithms, you aren’t just solving equations — you are building universes inside a machine.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b580db44e080" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Art of the Academic Cold Email: How to Secure a Postgraduate Research Advisor]]></title>
            <link>https://medium.com/@anilsarkar.phys/the-art-of-the-academic-cold-email-how-to-secure-a-postgraduate-research-advisor-7033cecb3567?source=rss-c9831424915d------2</link>
            <guid isPermaLink="false">https://medium.com/p/7033cecb3567</guid>
            <category><![CDATA[email]]></category>
            <category><![CDATA[cold-emails]]></category>
            <category><![CDATA[study-abroad]]></category>
            <category><![CDATA[physics]]></category>
            <category><![CDATA[research]]></category>
            <dc:creator><![CDATA[Anilsarkar Phys]]></dc:creator>
            <pubDate>Fri, 05 Jun 2026 03:14:07 GMT</pubDate>
            <atom:updated>2026-06-05T03:14:07.776Z</atom:updated>
            <content:encoded><![CDATA[<h4>A structured framework for prospective MSc and PhD students to stand out in a crowded inbox.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Dl1BuDXOoji3QFci" /><figcaption>Securing a place in a high-impact research group begins long before you step into the laboratory.. Source: sanjeri / Getty Images</figcaption></figure><p>Securing a post-graduate research opportunity is rarely just about meeting minimum university admission criteria. For prospective Master of Science (MSc) and Doctor of Philosophy (PhD) candidates, the true gateway to an academic programme is finding a faculty member willing to sponsor your research, fund your stipend, or supervise your thesis.</p><p>Top-tier professors receive dozens of generic, copy-pasted inquiries every single week. Most of these emails are archived within seconds because they fail to demonstrate a genuine understanding of the faculty member’s work. To stand out, your communication must be precise, analytical, and highly prioritised.</p><p>Here is a definitive strategy to transition your approach from cold outreach to a meaningful academic dialogue.</p><h3>Phase 1: The Groundwork of Alignment</h3><p>Before typing a single word of your email, you must conduct rigorous background research. You cannot convince a professor that you fit their laboratory if you only have a surface-level understanding of their discipline.</p><ul><li><strong>Analyse Recent Publications:</strong> Do not merely look at a professor’s university biography page; these are frequently outdated. Instead, look up their profile on Google Scholar or Scopus. Read their last three peer-reviewed papers published within the past two years.</li><li><strong>Identify the Current Trajectory:</strong> Pay close attention to the “Future Work” or “Discussion” sections of their recent papers. This reveals where their research group is moving next and allows you to position your skill set exactly where they need future labour.</li><li><strong>Check Lab Capacity:</strong> Visit the group’s dedicated website if they have one. Look at the current roster of PhD students and postdocs. If the website explicitly states “Not accepting new students for the upcoming autumn intake,” respect that boundary and pivot to another prospect.</li></ul><h3>Anatomy of a High-Conversion Academic Email</h3><p>An effective email to a professor is brief, formal, and structured logically. It should never read like a generic cover letter. Aim for a total length of 250 to 300 words, broken down into four distinct sections.</p><h3>1. The Subject Line (Clear and Contextual)</h3><p>Professors categorise their inbox rapidly. Your subject line must be explicit so they know exactly what you are requesting and what your credentials are.</p><ul><li><em>Weak:</em> Research Opportunity</li><li><em>Strong:</em> Prospective PhD Applicant: [Your University Name] Physics Graduate — [Specific Research Area]</li></ul><h3>2. The Hook (Who You Are and Why Them)</h3><p>State your name, your current academic status, and the exact purpose of your email immediately. Follow this by referencing a highly specific element of their work to prove you have done your homework.</p><blockquote><em>“Dear Professor [Surname],\n\nI am an honours graduate in Physics from National University, where I achieved a CGPA of 3.40. I am preparing an application for the upcoming MSc programme at your institution and am writing to express my interest in joining your research group, specifically focusing on your recent work regarding [insert specific mechanism or concept from their 2025/2026 paper].”</em></blockquote><h3>3. The Value Proposition (What You Bring)</h3><p>Connect your past academic achievements, laboratory skills, or thesis work directly to their ongoing projects. Highlight technical proficiencies, such as specific software, programming languages, or experimental techniques you have already mastered.</p><ul><li><strong>Highlight applicable methodologies:</strong> “My undergraduate thesis focused on [your topic], where I utilised [Method/Software]. This background aligns closely with your laboratory’s current investigations into [their topic].”</li><li><strong>Quantify your capabilities:</strong> Mention any relevant technical expertise, whether it involves computational modeling, data analysis pipelines, or complex instrumentation setups.</li></ul><h3>4. The Direct Ask (Call to Action)</h3><p>Conclude with a clear, low-friction request. Do not ask them to admit you on the spot; ask for a brief conversation to discuss mutual research alignments.</p><ul><li>“Are you expecting to accept new post-graduate students into your group for the upcoming academic cycle? If so, I would welcome the opportunity for a brief 10-minute video call to discuss how my background matches your current project requirements.”</li></ul><h3>What to Attach (And What to Omit)</h3><p>Your email must be self-contained but supported by evidence. Attach exactly two documents, cleanly formatted as PDFs:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1003/1*PhYu4SAjns1y2hdEwanHPQ.png" /></figure><p><em>Note: Do not attach a full Statement of Purpose (SOP) or a research proposal unless the professor’s website explicitly asks for it in initial inquiries. Keep the initial cognitive load light for the reader.</em></p><h3>Timing and the Protocol for Following Up</h3><p>Academic schedules dictate email responsiveness. Optimise your delivery times to match the professional habits of a university faculty member.</p><ul><li><strong>The Golden Window:</strong> Send your emails between 8:00 AM and 10:00 AM on a Tuesday, Wednesday, or Thursday in the professor’s local time zone. Avoid Monday mornings (when inboxes are flooded) and Friday afternoons (when attention shifts away from work).</li><li><strong>The Follow-Up Timeline:</strong> Professors are consistently overcommitted with lecturing, grading, and grant writing. If you do not receive a reply, wait <strong>7 to 10 full business days</strong> before following up.</li><li><strong>The Follow-Up Tone:</strong> When you do follow up, forward your original email with a polite, brief note at the top: <em>“Dear Professor [Surname], I am following up on my previous message regarding prospective MSc opportunities. I understand your schedule is incredibly busy, but I wanted to ensure my application materials reached you safely.”</em> If they do not reply to the second attempt, accept it as a passive refusal and redirect your energy toward other prospects.</li></ul><p>#AcademicWriting #GraduateSchool #ResearchOpportunities #PhDLife #HigherEducation</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7033cecb3567" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Best Python Libraries for Physics Students (And How to Actually Use Them)]]></title>
            <link>https://medium.com/@anilsarkar.phys/best-python-libraries-for-physics-students-and-how-to-actually-use-them-1174cceba645?source=rss-c9831424915d------2</link>
            <guid isPermaLink="false">https://medium.com/p/1174cceba645</guid>
            <category><![CDATA[physics]]></category>
            <category><![CDATA[simulation]]></category>
            <category><![CDATA[python-programming]]></category>
            <dc:creator><![CDATA[Anilsarkar Phys]]></dc:creator>
            <pubDate>Tue, 26 May 2026 12:14:28 GMT</pubDate>
            <atom:updated>2026-05-26T12:14:28.940Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/956/0*KJhUStUAqYOyt1lC.png" /></figure><p>*Stop fearing the code. Start falling in love with the physics.**</p><p>If you are a physics student in 2025, Python is no longer a “nice to have.” It is a fundamental tool, right up there with calculus and a good cup of coffee.</p><p>But here is the problem: Physics students are not software engineers. You don’t need to build a web app. You need to simulate projectile motion with air resistance, solve the Schrödinger equation, and fit a Gaussian to your messy lab data.</p><p>I’ve curated a list of the essential Python libraries for physics students. I’ve excluded the bloat and focused strictly on what moves the needle for your coursework and research.</p><p>Let’s dive in.</p><h3>1. NumPy (The Backbone)<br>Why: Pure Python is slow for physics. Really slow. NumPy gives you N-dimensional arrays and linear algebra at C-speed.</h3><p>Physics use case: Storing position/velocity vectors, matrix mechanics, Fourier transforms.<br><br><strong><em>import numpy as np<br># 1 million oscillations, instantly.<br>t = np.linspace(0, 10, 1000000)<br>x = np.cos(2 * np.pi * 5 * t) # 5 Hz wave</em></strong></p><h3>2. SciPy (The Workhorse)<br>Why: If NumPy is the engine, SciPy is the steering wheel. It contains actual scientific algorithms.</h3><p>The killer feature for physics students: scipy.integrate.solve_ivp (solving differential equations). Forget solving ODEs by hand for chaotic systems — let the computer do it.</p><p>Example: Solving a damped harmonic oscillator.<br><br><strong><em>from scipy.integrate import solve_ivp<br>import numpy as np</em></strong></p><p><strong><em>def damped_osc(t, y, b, w0):<br> x, v = y<br> dxdt = v<br> dvdt = -2*b*v — w0**2 * x<br> return [dxdt, dvdt]</em></strong></p><p><strong><em>sol = solve_ivp(damped_osc, [0, 10], [1, 0], args=(0.5, 2), t_eval=np.linspace(0,10,100))</em></strong></p><h3>3. Matplotlib (The Visualizer)<br>Why:You cannot see the physics in a list of numbers. You need plots. Matplotlib is the gold standard for publication-ready figures.</h3><p>Pro tip: Stop using the default settings. Use plt.style.use(seaborn-v0_8-darkgrid) for that premium look.</p><p>Physics use case: Plotting potential wells or wavefunctions.<br><br><strong><em>import matplotlib.pyplot as plt<br>x = np.linspace(-5, 5, 100)<br>V = x**2 # Quantum harmonic oscillator potential<br>plt.plot(x, V)<br>plt.fill_between(x, V, alpha=0.2) # Shade the well<br>plt.ylabel(‘Potential Energy’)<br>plt.show()</em></strong></p><h3><strong>4. SymPy (The Mathematician)<br></strong>Why: Sometimes you don’t want numbers; you want the symbolic answer. SymPy does algebra, calculus, and equation solving symbolically.</h3><p>Physics use case: Deriving the Euler-Lagrange equation for a Lagrangian without making an algebraic mistake.<br><br><strong><em>import sympy as sp<br>t, m, g = sp.symbols(‘t m g’)<br>x = sp.Function(‘x’)(t)<br>L = 0.5 * m * x.diff(t)**2 — m*g*x # Lagrangian T — V</em></strong></p><p><strong><em># Euler-Lagrange automatically<br>from sympy.calculus.euler import euler_equations<br>eq = euler_equations(L, x, t)<br>print(eq) # Output: -g — Derivative(x(t), t, t) == 0</em></strong></p><h3>5. lmfit (The Lab Savior)<br>Why: scipy.optimize.curve_fit is okay. `lmfit` is incredible. It is built specifically for physicists who need to fit models (Gaussians, exponentials, Lorentzians) to noisy experimental data.</h3><p>**Why students love it:** It gives you proper **error bars** on your fit parameters and handles bad initial guesses gracefully.</p><h3>6. QuTiP (The Quantum Mechanic)<br>**Why:** Quantum mechanics is hard. Matrices are big. QuTiP (Quantum Toolbox in Python) is designed for open quantum systems, spin chains, and cavity QED.</h3><p>**When to learn it:** If you are taking Advanced Quantum Mechanics or doing research in AMO physics.</p><p>**Example:** Creating a spin-1/2 state.<br><br><strong><em>import qutip as qt<br>up = qt.basis(2, 0) # |0&gt;<br>down = qt.basis(2, 1) # |1&gt;<br># Calculate expectation of sigma_x<br>sigma_x = qt.sigmax()<br>print(qt.expect(sigma_x, (up + down).unit()))</em></strong></p><h3>7. VPython (The 3D Classroom)<br>**Why:** Matplotlib is static. VPython runs in your browser or IDE and creates live, rotating 3D animations.</h3><p>**Physics use case:** Visualizing electric fields, orbital mechanics, or magnetic dipoles. It lowers the barrier to entry for computational physics dramatically.</p><p><strong><em>from vpython import *<br>sphere(pos=vector(0,0,0), radius=0.5, color=color.red) </em></strong><em># A proton</em><strong><em><br>arrow(pos=vector(1,0,0), axis=vector(1,0,0), color=color.green) </em></strong><em># A field</em></p><p># How to Build Your “Physics Stack”</p><p>You don’t need to install these one by one. Install **Anaconda** or **Miniconda**. It comes with NumPy, SciPy, Matplotlib, and SymPy pre-packaged.</p><p>```bash<br>conda install numpy scipy matplotlib sympy<br>pip install lmfit qutip vpython<br>```</p><p># The Golden Rule for Physics Students</p><p>**Don’t try to learn Python first. Learn Physics first, then translate it to Python.**</p><p>When you read a problem that says *”A particle moves under a central force”*, don’t open a Python textbook. Open a notebook. Write the equation. *Then* ask: “How do I turn this differential equation into a `solve_ivp` function?”</p><p>That is the mindset of a computational physicist.</p><p># Final Thoughts</p><p>You will hit bugs. Your plots will look weird. You will get `IndexError: list index out of range`. That is fine. Every professional physicist has been there.</p><p>The libraries above will handle the heavy lifting. Your job is to understand the physics well enough to tell the computer what to calculate.</p><p>**Which library are you most excited to try? Let me know in the comments.**</p><p>*Follow for more posts on Python for physicists, computational methods, and surviving grad school.*</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1174cceba645" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>