Coding classes in Mid and East Antrim

The learner starts with one line of reasoning: total gas equals the number of customers times the average used by each. So if customers grew by a factor of 242 divided by 171, about 1.42, and use per customer by 3, total gas grew by their product, about 4.25. Then comes the useful question: what steady yearly rate, compounded for 50 years, gives each factor? The answer is the factor raised to the power one fiftieth, minus one. Python makes this a two-line function, and the learner prints all three rates.

Two surprises come out. First, rates do not quite add: 0.70 plus 2.22 is 2.92, but the true total rate is 2.93, because the factors multiply rather than add. Over a few years the difference is invisible; over centuries it is not. Second, the share column uses logarithms, which turn multiplication into addition, so the growth splits cleanly: about 24 per cent from more customers and 76 per cent from each customer using more. The story in the record, new cookers, heaters and mantles, drove most of the rise, not new households.

The classic slip is to divide the whole rise by the years. The total went up by 325 per cent, and 325 divided by 50 is 6.5 per cent a year. But compound 6.5 per cent for 50 years and you get about 23 times the start, not 4.25. At the true rate of 2.93 per cent, gas use doubles about every 24 years. The start year matters a little too: counting 49 years instead of 50 nudges the total rate to 3.0 per cent, which is why careful code states its assumptions.

Running a quoted rate forward to test it is a reason for a Mid and East Antrim teenager to keep coding in 2026. The longer argument is in why teenagers should still learn to code in 2026.