Online coding and AI classes in the Puno region

Results published by MINEDU's assessment office (UMC) and reproduced in INEI's 2026 statistical compendium, first volume, as tables 5.53 (4° de primaria) and 5.54 (2° de secundaria).

MINEDU places a child by age on 31 March, under Law 28044's six years of primaria and five of secundaria, so a strong 10 year old in 5° de primaria can join an olympiad-style group with 11 and 12 year olds if the mentor agrees. Puno's public schools taught 93.9 thousand primaria pupils and 84.7 thousand secundaria students in 2025, with 27,116 teachers across all levels; the region's results come from their classrooms, and our role is only to add practice alongside them. For maths taught as its own subject in English, see our maths teachers in English for Peru.

import csv

def pct(text):
    return float(text.strip().rstrip("%"))

places = 0
covered = 0
with open("cobertura-movil-centro-poblado.csv", encoding="latin-1") as data:
    reader = csv.DictReader(data, delimiter=";")
    reader.fieldnames = [name.strip() for name in reader.fieldnames]
    first_4g = [name for name in reader.fieldnames if "_4G_CG" in name][0]
    for place in reader:
        if place["Departamento"] == "PUNO":
            places += 1
            if pct(place[first_4g]) >= 50:
                covered += 1

print("Column read:", first_4g)
print("Puno places in the file:", places)
print("Half or more under 4G:", covered)

Version 1 reads one operator and ignores three. By its count, Puno city is not covered, because 27% is under half. But a family in Puno only needs one network that works, so the right question for each place is: what is the highest figure among the four operators? That is a job for max() across the columns of the same row.

import csv

def pct(text):
    return float(text.strip().rstrip("%"))

places = {"URBANO": 0, "RURAL": 0}
covered = {"URBANO": 0, "RURAL": 0}
with open("cobertura-movil-centro-poblado.csv", encoding="latin-1") as data:
    reader = csv.DictReader(data, delimiter=";")
    reader.fieldnames = [name.strip() for name in reader.fieldnames]
    all_4g = [name for name in reader.fieldnames if name.endswith("_4G_CG")]
    for place in reader:
        if place["Departamento"] == "PUNO":
            kind = place["Clasificacion"]
            top = max(pct(place[name]) for name in all_4g)     # the strongest operator
            places[kind] += 1
            if top >= 50:
                covered[kind] += 1

print("Columns read:", all_4g)
for kind in places:
    print(kind, "places:", places[kind], "half or more under 4G:", covered[kind])
print("Total covered:", sum(covered.values()), "of", sum(places.values()))

The count rises from 299 to 457, and Puno city and Juliaca both count as covered. Yet only 65 of 164 urban places reach the line, which is surprising until you reread the dictionary: "CG" is the guaranteed footprint, a cautious figure. The file also gives "CG+CAR", which adds the operators' additional network capacity. Which one to use depends on the question. "Where can I rely on 4G?" suggests CG; "where is there 4G at all?" suggests CG+CAR.

import csv

def pct(text):
    return float(text.strip().rstrip("%"))

counts = {}
with open("cobertura-movil-centro-poblado.csv", encoding="latin-1") as data:
    reader = csv.DictReader(data, delimiter=";")
    reader.fieldnames = [name.strip() for name in reader.fieldnames]
    for footprint in ("_4G_CG", "_4G_CG+CAR"):
        counts[footprint] = 0
    columns = {fp: [n for n in reader.fieldnames if n.endswith(fp)] for fp in counts}
    for place in reader:
        if place["Departamento"] == "PUNO":
            for fp, names in columns.items():
                if max(pct(place[n]) for n in names) >= 50:
                    counts[fp] += 1

for fp, n in counts.items():
    print("Footprint", fp, "places half or more covered:", n)

With the wider footprint the count jumps more than elevenfold, to 5,304 of Puno's 10,544 places. None of the three numbers is wrong. They answer three different questions, and a careful report says which one it is answering. One more detail rewards a close reader: the dictionary lists the fourth operator's columns under a different company name from the one in the file header, a small reminder that the dictionary and the data can drift apart.

Data: OSIPTEL, Porcentaje de cobertura móvil por centro poblado, empresa operadora y tecnología, Plataforma Nacional de Datos Abiertos, licence ODC-By. Figures are declared by the operators and describe coverage footprints over each centro poblado's area, not measured speeds or the signal inside any home. The counts are ours; six provinces are left out of the table for space.

1. Under 9, with a map and coloured pencils. Four friends each shade the part of a school yard their torch can light. Is the yard lit if one torch covers it all and the others cover a little?
2. Scratch, around 9 to 11. Store four percentages in a list and show both the first one and the largest one. Which answers "can I get a signal?"
3. First Python, around 11 to 13. Try both of the first two programs on the Puno rows and explain why Puno city changes from "not covered" to "covered".
4. Ages 13 and over. Repeat Version 3 for 3G instead of 4G, then for your own province. Write one sentence stating exactly what each count measures.
5. Going further. Find the places where every column is 0%. Before writing "no signal", decide what the file can and cannot tell you about them.