Online coding and AI classes in the Huancavelica region

Figures from the household survey (ENAHO) as published in tables 11.5, 11.6 and 11.13 of the first volume of INEI's Compendio Estadístico Perú 2026; INEI writes the decimals with commas.

Read across the rows and the pattern is clear. In 2025 roughly a sixth of Huancavelica households owned a computer and fewer than a quarter paid for internet at home, while 69.1 per cent of residents from age six upward went online somewhere, most of them on a phone. Internet use has almost quadrupled since 2015. Home computers have not kept pace. That gap, not interest or ability, is the first thing a family here should think about before signing up for any online programming course, ours included.

From Cuadro N° 1 of the 2025 census tables (Censos Nacionales 2025), counting both people enumerated and people INEI estimates were missed, on the reference day of 15 September 2025. The department line is INEI's own total, not our addition.

Most of the region's people live far from any classroom that teaches programming. For a teenager in Churcampa or Castrovirreyna with a connection and a laptop at home, a live online class is one of very few ways to study coding with a person, rather than alone with videos.

import csv

home = total = 0
with open("ingresantes-unh-2025.csv", newline="", encoding="utf-8") as f:
    for row in csv.DictReader(f):
        people = int(row["TOTAL_INGRESANTES"])
        total += people
        if row["DEPA_PROCEDENCIA"] == "HUANCAVELICA":
            home += people

print("New students in 2025:", total)
print("From Huancavelica:", home)
print("Share:", round(100 * home / total, 1), "%")

Twenty-seven per cent suggests that most new students came from outside the region. That is not what the file shows. Version 1 treats every "NO INDICA" entrant as someone from elsewhere, when they could just as well be from Huancavelica. The only honest statement is a range: the lowest possible share assumes no unknown entrant is local, the highest assumes all of them are. Version 2 works that range out separately for each level of study, because the unknowns are not spread evenly.

import csv

counts = {}
with open("ingresantes-unh-2025.csv", newline="", encoding="utf-8") as f:
    for row in csv.DictReader(f):
        level = row["NIVEL"]
        origin = row["DEPA_PROCEDENCIA"]
        if origin == "HUANCAVELICA":
            kind = "home"
        elif origin == "NO INDICA":
            kind = "unknown"
        else:
            kind = "elsewhere"
        tally = counts.setdefault(level, {"home": 0, "unknown": 0, "elsewhere": 0})
        tally[kind] += int(row["TOTAL_INGRESANTES"])

for level, t in sorted(counts.items()):
    total = sum(t.values())
    low = 100 * t["home"] / total
    high = 100 * (t["home"] + t["unknown"]) / total
    print(level, "| total", total, "| unknown", t["unknown"],
          "| between", round(low, 1), "and", round(high, 1), "%")

Data: Universidad Nacional de Huancavelica, Lista de Ingresantes del año 2025 I y II a la UNH, ODC-By. People are the sum of TOTAL_INGRESANTES; shares and bounds are ours. The file has 1,417 rows standing for 2,010 entrants.

Now the answer has two very different parts. For undergraduates the range is narrow, about 90 to 94 per cent, because only 17 of 412 did not state an origin: the university's first-degree students are overwhelmingly from the region. For postgraduate and second-speciality entrants the range runs from almost nothing to nearly nine in ten, which means the file simply cannot answer the question for them. Saying "we cannot tell" is a result, and a more useful one than 26.8 per cent.

1. Around 8, with counters. Ten counters in a bag: four red, two blue, four hidden under a cloth. How many red could there be? Write the smallest and the largest.
2. Blocks, around 9 to 11. Make a list of ten answers with three marked "?". Count the yes answers twice, once treating "?" as no and once as yes.
3. First Python, around 11 to 13. Edit Version 1 so that it counts JUNIN. Is the new share a lower bound, an upper bound or neither?
4. Ages 13 and over. Add a third line to Version 2 that prints the share among entrants who did state an origin. Explain why it sits between the two bounds.
5. Going further. Group the undergraduate rows by ESPECIALIDAD and find the programmes whose range is widest. Which ones should the university ask about first?

Suggestions of ours, not rules. Because MINEDU places pupils by their age on 31 March, two pupils in one grade may differ by nearly twelve months.

Teenagers will find more on our teen coding page for Peru, and engineering students may prefer Python for engineering students. The Peru hub page lists every guide we have written for the country.