Online coding and AI classes in the Junín region

INEI, Censos Nacionales 2025, Cuadro N° 1 of the demographic indicators: total population at 15 September 2025, within boundaries as of 15 July 2026. Each figure is INEI's; nothing has been added up by us. "Nearly half" is our rounding of 626,199 out of 1,422,780.

From Ley 28044 (six years of primaria, five of secundaria), MINEDU's normative ages counted on 31 March, and the CNEB secundaria plan of studies. Read in November 2026.

import csv

days = {}
with open("bsrn-2025.csv", newline="", encoding="utf-8") as station:
    for row in csv.DictReader(station):
        key = (int(row["month"]), int(row["day"]))
        days.setdefault(key, [None] * 24)
        if row["global_SW_rad"] != "":
            days[key][int(row["hour"])] = float(row["global_SW_rad"])

starts = {}
complete = 0
for hours in days.values():
    if None in hours:
        continue
    complete += 1
    peak = max(range(22), key=lambda s: sum(hours[s:s + 3]))
    starts[peak] = starts.get(peak, 0) + 1

print("Days in the file:", len(days))
print("Days with no gap:", complete)
print("Peak three hours start at:", max(starts, key=starts.get))

Version 1 works, but look at what it threw away. It only uses days with no gap at all, and because a single blank hour, even at night, ruins a day, only 92 of 365 days survive. It also adds up every three-hour slice from scratch, which is fine here and wasteful on longer data. The answer, a window starting with the reading labelled 11, is reasonable, but it rests on a quarter of the year.

Version 2 slides a three-hour window along each day. It keeps a running total and a count of gaps: when an hour enters the window, it is added (or the gap count goes up); when an hour leaves, it is taken away (or the gap count goes down). A window is only considered when its gap count is zero, so a blank hour at midnight no longer spoils the afternoon. It also flags any day whose peak window starts before six in the morning, because that cannot be right for sunlight.

import csv

days = {}
with open("bsrn-2025.csv", newline="", encoding="utf-8") as station:
    for row in csv.DictReader(station):
        key = (int(row["month"]), int(row["day"]))
        days.setdefault(key, [None] * 24)
        if row["global_SW_rad"] != "":
            days[key][int(row["hour"])] = float(row["global_SW_rad"])

def best_window(hours, width=3):
    total, gaps, peak = 0.0, 0, None
    for h in range(24):
        if hours[h] is None:
            gaps += 1
        else:
            total += hours[h]
        if h >= width:
            leaving = hours[h - width]
            if leaving is None:
                gaps -= 1
            else:
                total -= leaving
        if h >= width - 1 and gaps == 0:
            if peak is None or total > peak[0]:
                peak = (total, h - width + 1)
    return peak

starts = {}
odd = []
for key, hours in days.items():
    found = best_window(hours)
    if found is None or found[0] < 1:   # night, or no daylight in any gap-free window
        continue
    starts[found[1]] = starts.get(found[1], 0) + 1
    if found[1] < 6:
        odd.append(key)

print("Days we can use:", sum(starts.values()))
print("Most common starts:", sorted(starts.items(), key=lambda p: -p[1])[:3])
print("Check by hand (month, day):", odd)

Now 317 days count instead of 92, and the answer holds: on 139 of them the sunniest three hours begin with the reading labelled 11, and on most of the rest they begin at 10 or 9. The check catches one day, 21 October 2025. Opening it shows why: that day has daylight readings only at 6, 12, 13, 15 and 16, so every window covering the middle of the day contains a gap, and the "peak" complete window is in the early morning. The program did exactly what it was told; the flag makes sure a person looks before anyone trusts that day. One more detail for the curious: running totals of decimals can leave a tiny leftover such as 0.00000000000003 instead of zero, which is why Version 2 ignores totals under 1.

Data: Instituto Geofísico del Perú (IGP), LAMAR, Estación BSRN para el Balance Energético Solar Tierra-Atmósfera de LAMAR - Junín, Perú, 2025 file, ODC-By. Averages are ours, over the hours that have a reading; the data dictionary describes the hour column as the average hour of the record. Blank readings by month: January 26, February 97, March 744 (the whole month), April 222, May 7, June 12, July 110, August 85, September 53, October 114, November 143, December 71.

1. Under 8, outdoors. Put a stick in the ground and mark the end of its shadow three times in a morning. When is the shadow shortest? What does that tell you about the sun?
2. Blocks, around 8 to 10. Make a list of twelve numbers for the hours of daylight and find the biggest single number. Then find the biggest sum of two numbers side by side.
3. First Python, around 10 to 12. Run Version 1. Change the window from three hours to two, then to four. Does the peak starting hour move?
4. Ages 13 and over. In Version 2, print the window totals for 21 October 2025 hour by hour and explain, in two sentences, why the peak gap-free window is so early.
5. Going further. Group the 317 usable days by month and find, for each month, the most common starting reading. Is the pattern the same in the rainy and the dry months?

INEI, Compendio Estadístico Perú 2026, tomo 1, cuadros 11.5, 11.6 and 11.13 (Encuesta Nacional de Hogares). INEI uses decimal commas; we print points.

Pairings are our suggestion. Because MINEDU fixes the age on 31 March, children born after that date may be a grade lower than shown.

Parents of younger children may like Python for kids in Peru, and Spanish speakers programación para niños de 12 años. University students in engineering can read Python for engineering students in Peru, and the national overview is coding and AI classes in Peru.

Peru keeps UTC minus 5 hours with no daylight saving (tz database, zone America/Lima, unchanged since 1994). India keeps UTC plus 5 hours 30 minutes. The gap is 10 hours 30 minutes every day of the year.

With no clock changes on either side, a class hour stays the same from one season to the next. You pick a time in your own clock when booking. The 2026 school year in Peru opened on 16 March 2026, and families often start in the long summer holiday and keep the hour when school resumes.