Physics and Python glossary
Use these definitions while reading the lessons or running a notebook. Each term includes a practical example or a check you can apply to your own calculation. Start with Python, Jupyter notebook, cell, and runtime if you are new to Google Colab.
- assumption
- A statement we accept for a particular model. For a small lift near Earth, we may treat gravity as constant. This makes the calculation simpler, but it also limits where the answer is useful.
- boundary condition
- A rule at the edge of a region. A string fixed at both ends has zero displacement at those ends. Boundary conditions describe the edges; initial conditions describe the starting time.
- conservation law
- A statement that a quantity stays constant when the relevant system is isolated. Total energy can stay constant while kinetic energy becomes potential energy. Checking their sum helps reveal mistakes in a simulation.
- coordinate system
- The origin and directions used to describe positions. A negative position means an object lies on the negative side of the chosen origin. Changing coordinates changes the numbers, while the physical situation stays the same.
- differential equation
- An equation involving a quantity and its rate of change. Velocity is the rate of change of position. Given an equation and starting values, a numerical method can estimate the motion in small time steps.
- field
- A quantity defined at positions throughout a region. Temperature is a scalar field: each location has a temperature. An electric field also has direction, so its value needs vector components.
- invariant
- A property that remains unchanged during a specified process or transformation. Rotating coordinate axes preserves the length of a vector. A useful check computes that length before and after the rotation.
- model
- A limited description used to answer a physical question. A projectile model may include gravity and omit air resistance. State what is included, what is omitted, and what conditions make those choices reasonable.
- numerical error
- A difference introduced by finite arithmetic or an approximate numerical method. Reducing a time step can reduce some errors, but rounding error does not always shrink with it. This differs from uncertainty in a measurement.
- observable
- A quantity that can be compared with a measurement, such as position, temperature, or a detector count. A model may contain other internal quantities; explain how its calculated values connect to the observable you want.
- parameter
- A value held fixed during one calculation, such as a spring constant. You can change it between runs to compare systems. Distinguish these choices from quantities, such as position, that evolve during a run.
- state
- The information needed to describe the system at a given time and predict its evolution under the model. For a particle obeying Newton’s laws, this usually includes position and velocity, along with the specified forces.
- unit
- An agreed size for a quantity, such as metre, second, or joule. Convert centimetres to metres before combining them with values in metres. A variable name ending in _m is a reminder, not automatic unit checking.
- validation
- Checking whether a model answers its intended physical question well enough. Compare predictions with reliable measurements and established results. Separately check the code against known solutions; correctly implemented equations can still describe an unsuitable model.
- vector
- A quantity with magnitude and direction. Its components depend on the chosen axes. A NumPy array can store those components, but an array of temperatures at different times is not automatically a physical vector.
- python
- The programming language used for the course examples. An assignment such as mass_kg = 2.0 stores a value under a name. Python runs the instructions you write; it does not infer missing units or physical assumptions.
- jupyter notebook
- A document containing text cells, executable code cells, and their outputs. Save it as an .ipynb file. The visible order of cells should reproduce the result when you restart Python and run from the top.
- google colab
- A hosted notebook service that runs Python in a browser. Open a course notebook, sign in to run code, and save a copy in Drive to keep changes. Runtime memory and temporary files do not last indefinitely.
- cell
- One block in a notebook. A text cell explains a question or result; a code cell executes Python. Editing an earlier cell does not automatically update later outputs: rerun the cells that depend on it.
- runtime
- The running Python environment that holds variables, imported libraries, and temporary files. Restarting it clears that working memory. A saved notebook can contain old outputs even when its current runtime has no variables.
- numpy
- A Python library for numerical arrays. Use it to calculate positions at many times, summarize samples, or perform matrix operations. Check array shapes and units: successful arithmetic alone does not establish a correct physical calculation.
- sympy
- A Python library for symbolic mathematics. It can rearrange equations, differentiate expressions, and substitute values. A symbol represents a mathematical unknown; assumptions about its sign or domain affect which answers are physically meaningful.
- matplotlib
- A Python plotting library. Label axes with quantities and units, and identify whether points are measurements or model predictions. More plotted points can reveal a curve more clearly, but they do not improve the underlying measurements.
- initial condition
- A value specified at the starting time, such as a particle’s initial position and velocity. Use the same time origin as the equation. Changing an initial condition selects a different solution of the same motion law.
- uncertainty
- A description of how well a measured or inferred value is known. Record what a quoted range or standard uncertainty means. Extra decimal places from Python do not create more precise knowledge of the physical input.