An algorithm describes precise steps for solving a problem. A program expresses those steps in an executable language. Decomposition separates tasks such as receiving a measurement, interpreting it and communicating an alert. Abstraction keeps the information needed and leaves out details that do not affect that decision. Before choosing blocks or writing instructions, we must agree on the input and the result we expect.

Input, processing and output describe the path of information. A measurement needs meaning, a range and a scale or unit: the number 24 could represent different things. In this course we can simulate light from 0 to 100 without physical units; this scale is not lux. Typing a number provides a test input, not a sensor reading. A rule written by a person can already automate a response; it does not imply machine learning. Scratch and Python can preserve the algorithm even when their notation and screens differ.

The editor lets us write code; the Python interpreter executes instructions; the .py file preserves the program. Thonny combines an editor and console, where results appear and questions are answered. Saving preserves program text, but does not automatically save answers from a previous run. We therefore need to recognise the open file, save it in a working folder and run it again after editing.

In a simple initial sequence, instructions run from top to bottom. print() displays information and receives its data inside parentheses; text is enclosed in quotation marks. Python is case sensitive, so print and Print are different names. The >>> symbol in an interactive session is a console prompt and must not be copied into the file. A comment starts with # outside a string and explains intention or limitations; it does not perform the action described. Later, conditions, loops and calls will change the program’s execution path.

A variable keeps a value under a name so it can be used later or replaced when another input arrives. The name expresses what the value represents in the problem. In nivel = 18, = performs assignment; nivel == 18 asks whether the values are equal. When executing contador = contador + 1, Python first evaluates the right-hand side using the previous value and then assigns the result. This is why an update makes sense even though the same name appears on both sides. Clear names without spaces or an initial digit help us follow these changes.

The type matters as well as the value: int represents integers, float numbers with decimals, str text and bool Boolean values. Operations depend on this distinction. Joining strings containing digits preserves a sequence of characters; adding their numeric values performs a calculation. Since input() returns text, an answer that looks like a number still needs conversion before it can be used as one. float() attempts this step; storing the answer in a variable does not do it automatically.

Conversion and validity therefore answer different questions. In these exercises, the agreed format uses a decimal point and leaves the unit out of the answer. Letters or a decimal comma can prevent conversion. A negative value, however, can convert successfully while falling outside the problem’s permitted range. The basic program assumes valid numbers; recovering from incorrect input remains an explicit extension. This distinction lets us explain each outcome without confusing a formatting problem with a measurement outside the scale.

A condition produces a Boolean result: True or False. The operators <, >, <=, >=, == and != express less than, greater than, inclusion of equality, equality and inequality. A threshold separates behaviours; deciding what happens exactly at it belongs to the design. A strict condition does not include equality. Predict that case before running, because many errors appear only at the boundary.

if and else choose one of two paths. Python puts a colon at the end of the header and uses indentation to delimit its body; we will use four spaces per level. An instruction aligned outside both bodies runs after the decision. Two independent if statements can trigger two responses, whereas if/else expresses exclusive alternatives. As an extension, and requires both conditions, or accepts at least one and not reverses a result. Checking a range with and requires a number first; it does not itself repair a failed text conversion.

A loop repeats a body of instructions. An iteration is one complete pass. In Python, for turno in range(5) visits 0, 1, 2, 3 and 4: the final bound is excluded and five iterations run. range(1, 6) also has five values but starts at 1. Displaying turno + 1 changes the numbering presented, not the values generated by range.

The position of each instruction determines what repeats. If input() is inside the body, an answer is requested on every iteration; outside, the previous value is reused. The same applies to a condition and a function call. A closing message placed outside the loop appears once. Scratch can express a counted repetition with blocks, although it does not necessarily expose the same counter. Five queries provide a clear endpoint for comparing programs. A continuously observing device also needs decisions about frequency and stopping; replacing five with a huge number is insufficient.

A function groups the instructions for a particular task under a name. This organisation allows the task to be checked separately and used later at different points in the program. Defining the function prepares its body but does not start execution. A call directs execution to those instructions; when they finish, the program continues after the call. Separating definition from calling explains why a program can contain a correctly written function without producing any of the messages we expect from it.

Parameters allow the task to receive data without requiring a different definition for every value. In def mostrar(valor), valor is the name used inside the function: the parameter. In mostrar(17), the argument is the specific value, 17, received by that call. The same definition can handle other calls with different arguments. In the basic monitor, the function communicates the alert corresponding to the received measurement and is called on every loop iteration. This connects the new input, the task and its outcome.

A Scratch custom block with an input expresses an organisation comparable to def and its indented body in Python. There are also correspondences between the green flag and running the file, asking a question and receiving information, variables, conditions and repetition. Comparing behaviour helps preserve the algorithm, although differences in types and some operations prevent automatic translation. Within Python, print() displays information and return passes a result to another part of the program. pass allows a valid body that does not yet perform the intended task; its presence does not establish that the task has been completed.

A trace records the actual execution path: input, variable, received parameter, comparison, Boolean result and output. It separates what the code does from what the design says it should do. Prepare a prediction first and then compare it with execution. If they differ, either the program or the prediction may be wrong. Following a complete iteration avoids attributing an output to a previous input.

Debugging investigates a discrepancy and corrects its cause. A syntax error prevents a structure from being interpreted; a runtime error occurs during an operation, such as converting letters; a logic error can finish without warnings while making the wrong decision. The message and indicated line provide clues, although the cause may be earlier. Reproduce the symptom, propose a hypothesis, change one cause and repeat the same test. Then check other previously working cases. Record the sequence, expected and observed results, cause, repair and rerun, not just a screenshot of corrected code.

A test relates input data to expected results defined by the design. For a boundary, use a value below it, the threshold itself and a value above it; for execution flow, alternate inputs and repeat one. The ends of the permitted range and the threshold check different properties. A decimal examines numeric handling. A nonnumeric input explores a limitation of the basic version and must not be presented as a valid case successfully handled.

Comparing Scratch and Python requires the same data in the same order. Their messages may look different while communicating equivalent decisions. If both agree on an incorrect response, they still violate the design: compare them with the intended rule too. In the project, four complete runs of five inputs combine boundary, decimal and repetition tests. The record preserves two incidents and one verified improvement; introduce controlled faults only in a copy. Passing these cases provides specific evidence without proving that every imaginable input will work.