What are the best coding classes in Rotterdam Centrum?
Coding classes in Rotterdam Centrum reach 42,924 residents (1 January 2026) on 413 hectares of land, the smallest of the fourteen gebieden, in 25,035 households of which 15,290 hold one person. Six primary establishments here teach 1,989 pupils and four secondary ones teach 1,622, in a district holding 3,130 residents under fifteen. None of those figures answers the question anybody wants answered, because a school catchment is not district-shaped and this district is small enough that almost everyone in it lives near an edge. Modern Age Coders teaches live online in English across all six buurten, ages 6 to 67, first lesson free, then USD 100 monthly in a group of five to ten or USD 150 one to one.
Rotterdam Centrum covers 413 hectares of land, which is 4.13 square kilometres. Lay that out as a square and its sides are about two kilometres, so a person standing anywhere in it is on average a few minutes from a district that is not this one. Count what is near residents by counting only what falls inside the boundary and you will undercount for nearly all of them, not because the centre is badly served but because it is small. The same count in a district four times the size undercounts far less. Half of what separates those two numbers is geometry.
Facts last verified 9 September 2026. Teaching is online; no Rotterdam Centrum branch is claimed.
A groep 5 child in the Oude Westen who has finished every game in the house, a teenager off the Coolsingel who wants to see what a model does rather than what it says, a student near Dijkzigt who needs Python to be automatic, and an adult in the Stadsdriehoek who has to defend a figure in a meeting every month: four doors, and the first hour of each is free.

Real typed Python, drawings that take instructions and a first program that works out a rule, taught in English to the groep 5 to 8 child in any of the six buurten.
See the syllabus
A model trained, tested and deliberately broken, for the teenager who has used plenty and been shown the inside of none.
See the syllabus
Python that collects, cleans, checks and reports on its own schedule, for the student or worker doing one job by hand every week.
See the syllabus
Cleaning, joining, and knowing what a boundary did to a number before you quote it, for anybody who hands figures to people who act on them.
See the syllabusBrowse the course atlas for more than one hundred options and use the coding roadmap to check prerequisites.
The four we are known for
These run underneath everything above. Every one is live and online, placed by ability rather than by age, and the first class is free.

Python, web and AI projects where the learner still owns the thinking.
See the syllabus
Automate the work you already do, then let AI carry part of it.
See the syllabus
Train a model, read what it learned, and be able to say why it is wrong.
See the syllabus
Run AI coding agents on real work without losing control of the codebase.
See the syllabusRotterdam counts 673,804 residents across 22 wijken and 92 buurten. Its centre holds 42,924 of them on 488 hectares, of which 413 are land and 75 are water, at a published 10,191 people to the square kilometre. It is the smallest of the fourteen gebieden by population and the one the rest of the city passes through, and the city page in this series carries the port and the school-choice picture so that this one does not have to.
The centre is divided into the Cs Kwartier, the Stadsdriehoek, Cool, the Oude Westen, Dijkzigt and the Nieuwe Werk. 3,130 of its residents are under fifteen, 7.3 percent, the lowest share on any page in this series so far. 18,750 are between twenty-five and forty-five, 44 percent, and 15,290 of its 25,035 households hold exactly one person, 61 percent.
23,478 dwellings, of which 17,139 are rented, 73 percent, at an average value of 424,000 euro. Average income is 44,500 euro per resident for 2024, which is high for Rotterdam and unsurprising in a district whose population is overwhelmingly of working age.
Six primary establishments here teach 1,989 pupils and four secondary establishments teach 1,622. Those are counts of pupils attending schools inside the boundary, not of children living inside it, and in a district this small the two have very little to do with each other.
Where Rotterdam differs from Amsterdam, for anyone reading both
The national statistics office publishes Rotterdam at wijk level with the wijken corresponding to the gebieden, so this district has a full statistics row of its own. Amsterdam does not work that way: its stadsdelen are not a publication level at all, which is why the Amsterdam district pages in this series are built from the wijken beneath them instead. One consequence is that a Rotterdam wijk holds about 30,600 people on average and an Amsterdam wijk about 8,500. What that does to any national table with a wijk column is the argument of the Tilburg page and is not repeated here.
Prins Alexander is the largest of the fourteen and sits on the north-eastern edge of the city. Setting it beside the centre shows what a boundary does to a number, because the two are described by identical columns and are nothing like the same shape.
| Rotterdam Centrum | Prins Alexander | What the difference is made of | |
|---|---|---|---|
| Residents | 42,924 | 97,794 | the smallest gebied and the largest |
| Land area | 413 hectares | 1,738 hectares | four times as much ground |
| People per square kilometre | 10,191 | 5,635 | the centre is nearly twice as dense |
| Residents under fifteen | 3,130, 7.3 percent | 14,585, 15 percent | a suburb has children and a centre does not |
| Primary establishments, and pupils | 6, teaching 1,989 | 28, teaching 8,108 | 332 pupils each against 290 |
| Secondary establishments, and pupils | 4, teaching 1,622 | 9, teaching 5,860 | 406 pupils each against 651 |
| Households, and average size | 25,035 at 1.6 | 48,260 at 2.0 | 61 percent single-person against 45 |
| Share of the area more than 500 m from an edge | about 26 percent | about 58 percent | geometry, not provision |
That last row is this page modelling each district as a square of its published land area and asking how much of it lies more than five hundred metres inside the boundary. Neither district is a square, and real shapes are more irregular than squares, which makes the edge share larger rather than smaller. The figure is an illustration of a mechanism rather than a measurement of these two places, and it is labelled as one wherever it appears. What it shows is that the centre is mostly edge and the suburb is mostly interior, before anybody has counted a single school.
One city total, two district rows, and a pair of calculations this page performs deliberately in order to show what they are worth.
42,924 residents (2026); 25,035 households averaging 1.6 (2025), 15,290 single-person at 61 percent; ages 3,130 under fifteen at 7.3 percent, 7,455 at 17 percent, 18,750 at 44 percent, 8,460 at 20 percent and 5,115 at 12 percent; 488 hectares total, 413 land and 75 water; density 10,191; 23,478 dwellings; average value 424,000 euro; 17,139 rented at 73 percent and 6,339 owned at 27; income 44,500 euro (2024); six buurten; six primary establishments with 1,989 pupils and four secondary with 1,622.
97,794 residents (2026); 48,260 households averaging 2.0 (2025), 21,710 single-person at 45 percent; 14,585 residents under fifteen at 15 percent; 1,860 hectares total, 1,738 land and 122 water; density 5,635; 47,041 dwellings; average value 360,000 euro; 26,343 rented at 56 percent and 20,698 owned at 44; income 34,700 euro (2024); eight buurten, among them Ommoord, Zevenkamp and Nesselande; 28 primary establishments with 8,108 pupils and nine secondary with 5,860.
Rotterdam: 673,804 residents in 2026, 342,340 households in 2025, 22 wijken and 92 buurten. The city page in this series carries the port throughput argument and the secondary-school picture, and neither is repeated here.
Divide primary pupils by residents under fifteen and the centre gives 0.64 and Prins Alexander 0.56. Do the same with secondary pupils and you get 0.52 and 0.40. Both calculations divide a count of people attending schools inside a boundary by a count of people living inside it. Those are different populations, the ratio is meaningless in both districts, and it is printed here only because seeing the number makes the objection concrete.
This page does not say how well served children in the centre are, because no figure read here answers that. It does not know where the 1,989 primary pupils live, where the district's 3,130 children go to school, or how far any of them travel. Catchment data would answer those questions and none was found.
Modern Age Coders has no connection to the gebied, the city or any school named here and claims none. The square-equivalent sides of about 2.03 and 4.17 kilometres, the edge shares of about 26 and 58 percent, the pupils-per-establishment figures of 332, 290, 406 and 651 and the two ratios above are all this page computing from published numbers, and every one of them is labelled as arithmetic where it appears.
A boundary, a radius, and a bias that grows as the area shrinks without anything on the ground changing at all.
Ask what is within five hundred metres of a resident and you get a circle. Ask the same question of a dataset organised by district and you get the part of that circle that falls inside the boundary. For somebody in the middle of a large district those are the same thing. For somebody near an edge, most of the circle is missing, and the data has no way to say so.
Model the centre as a square of its 413 hectares and its sides are about 2.03 kilometres, so roughly 26 percent of it sits more than five hundred metres from an edge. Do the same for Prins Alexander at 1,738 hectares and the figure is about 58 percent. Same rule, same radius, and more than twice the interior, purely because the second district is bigger.
Any table of districts ordered by amenities near residents, counted within boundaries, is partly a table ordered by district size and shape. Small districts look badly served and large ones look well served, and a good deal of that ordering would survive even if every district had identical provision on the ground.
| The question | How it gets computed | What the boundary does to it | A better approach |
|---|---|---|---|
| How many schools are near residents here? | count schools inside the district | cuts off every school just over the line | count within a radius of each resident, ignoring boundaries |
| How many pupils does this district educate? | sum enrolment at schools inside it | counts pupils who live elsewhere | ask for pupil home postcodes, if they exist |
| How many children live here? | the resident count, 3,130 under fifteen | nothing, this one is fine | use it, and do not divide it by the previous row |
| Is the centre well served? | a ratio of the first two | produces 0.64, which describes nobody | accept that the published data does not answer it |
Almost every dataset arrives cut into regions, and almost every question about proximity gets answered inside those cuts because that is how the data is shaped. The trouble is that people and the things they use do not respect the cuts. A resident near a border has half their world in the next region, the count inside the border misses it, and nothing in the output records the omission. What makes this worse than an ordinary inaccuracy is that the size of the error depends on the size of the region. A small region is nearly all edge; a large one is mostly interior. So when you line regions up and compare them on any within-boundary measure of what is nearby, part of what you have measured is how big each region is. The same failure runs through sales territories, hospital trusts, police areas, delivery zones, store catchments, network regions and school districts. In each case the honest fix is to stop letting the boundary define the question: compute proximity from each point outward, in whatever radius the thing actually reaches, and let the results cross the lines the way the people do. When that is impossible because the data only exists per region, the correct output is not a ranked table. It is a note saying that the measure is biased by area size, and that small regions will look worse whether or not they are.
Sales territories, delivery zones, hospital catchments, store radii, network regions: the boundary was drawn for administration and the question is about distance. Practised on a district that is mostly edge.
| Stage | Rule | What it prevents | What it gives you |
|---|---|---|---|
| Separate the two questions | Ask whether you want what is inside the region or what is near the people | A proximity question answered by a container | Clarity about what the number can mean |
| Measure the edge share | Estimate how much of each region lies within the relevant radius of a border | Regions compared as though the bias were equal | A size for the distortion, per region |
| Compute outward, not inward | Where the coordinates exist, measure from each point rather than per region | Everything just over the line discarded | Numbers that describe people rather than shapes |
| Never rank on a truncated measure | If only regional counts exist, refuse the league table | Small regions penalised by geometry | An honest gap where a ranking would have been |
| Say which population a count is of | Distinguish who is inside from who uses the thing | Attendance divided by residence | Two useful numbers instead of one useless ratio |
Draw a circle around your house on a map of the playground, then draw the line where the playground ends. The children count what is inside both, then count what is really within reach, and discover the two answers differ most for whoever sits closest to the fence.
The two districts in Python, a function that computes the interior share of a square for a given radius, and a comparison that prints that share beside every per-district count. The stretch: produce the ranked table of districts by schools per child, then write the sentence explaining why the ranking is partly a ranking by area.
A real regional measure from work, such as coverage, response time or store proximity. Adults reliably find the regions were drawn for reporting, the question is about distance, and the smallest region has been on an improvement plan for two years.
All population, household, age, area, density, dwelling, value, tenure, income, establishment and pupil figures for both districts, and the city totals, come from the national statistics office pages named beside them, with 2026 populations and pupil counts, 2025 household and housing figures and 2024 incomes. The square-equivalent sides, the edge shares of about 26 and 58 percent, the pupils per establishment and the two attendance-over-residence ratios are this page computing from those figures, and the square model is stated as a model rather than as a description of either district.
The rung is settled in the free hour by what the learner can already do, and never by which of the six buurten the address falls in.
Children draw a circle, meet a fence, and count twice.
Coding for KidsScratch Coding for KidsLearners work out which of the two questions a count has actually answered.
Python and AI for KidsMaths Through CodingTeenagers compute how much of an area is edge and print it beside every regional figure.
AI and Machine Learning for TeensData Science for TeensAdults take a regional coverage measure from work and find the geometry inside it.
Python Automation CourseData Analysis CourseBecause part of that answer was decided by where somebody drew a line, and none of it is visible in the numbers.
Give a model the counts and it will produce a comparison, and the comparison will be about provision: how many schools, how many children, which district comes off worse. Every figure it uses is correct and the arithmetic is sound. What is missing is that one district is 413 hectares and the other 1,738, that the smaller one is mostly within a few minutes of somewhere else, and that a count made inside a boundary discards whatever sits just over it. Nothing in a table of counts carries the shape of the thing being counted.
So the learner practises asking what the boundary did before asking what the number says, and then builds the version that measures outward from people instead of inward from lines. A fourteen-year-old off the Coolsingel who has worked out why the small district looks badly served will be the person at thirty asking whether the underperforming sales region is underperforming or merely small. Models will keep comparing the columns they are given, fluently and correctly. Noticing what the columns left outside stays a person's job.
The case for a child in Rotterdam Centrum learning to code in 2026 is therefore not about schools. It is that a great deal of working life is spent comparing regions that were drawn for administrative convenience, that any question about distance answered inside those lines is biased by their size, and that a district three quarters of which is within five hundred metres of somewhere else makes the bias impossible to ignore. The longer argument is in why a child should still learn to code in 2026.
Nothing in the centre is far from anything. What families here run short of is space and evenings, in flats where 61 percent of households are one person and the rest are fitting a great deal into small rooms.
The lesson happens at the table in the Stadsdriehoek, the Oude Westen or Cool, at one hour that repeats every week, which is the only arrangement that does not need a spare room somewhere else.
Teaching is in English throughout. Groep, havo, vwo and profielkeuze stay Dutch, because those four are printed on the paperwork a child brings home and an English version would only need turning back.
A level, found by giving the learner something real to do. A course chosen because that level points at it. A weekly time. Nothing about payment comes up while any of it is happening.
Five to ten learners at one level. In a district with 3,130 children under fifteen, drawing from further afield is not a preference, it is the only way a specific level fills.
Two lessons weekly at a repeating hour, normally eight in a month. Zuid-Holland takes its school holidays with regio Midden, and the weeks a family will be away are agreed at the start rather than found later.
The Dutch clock changes twice a year and the Indian one does not, which is where the whole difference comes from. Late afternoon, early evening and either weekend day work at both settings.
The lowest share of children of any page in this series
Just 3,130 of the centre's 42,924 residents are under fifteen, 7.3 percent, against 15 percent in Prins Alexander and 19 percent in the fullest Amsterdam neighbourhood this series has measured. A child growing up here is in a very small local cohort, and no school or club inside four square kilometres can assemble a group of five to ten at one particular level in one particular subject from it. That is arithmetic rather than any comment on the schools, and it is the reason groups here are formed on ability from across the world, at every age up to sixty-seven.
On the page rather than saved for a telephone call.
A complete lesson with a real task, a level and a named next course.
Per month, normally eight lessons, five to ten learners placed by ability.
Per month, normally eight lessons, one learner with one teacher.
One international rate applies and it is quoted in dollars, with no euro edition, so an address in Dijkzigt is charged exactly what one in the Cs Kwartier is. Nothing is paid until the free lesson has settled a course and an hour, and payment is arranged over WhatsApp afterwards. Pausing, switching plan and missing a week are described on the pricing page.
Rated 4.9 across 547 Google reviews. These are real reviews, reproduced as written.
★★★★★
"The one step solution for my son. Modern Age Coders make learning coding so simple that kids love it. The teachers explain complex concepts clearly with practical exercises and interactive content."
Ria Mukherjee
Parent
★★★★★
"Modern Age Coders has been a game-changer for me. I struggled to grasp IT concepts and coding before joining, but their classes transformed everything. I can now confidently write complex programs with ease."
Samriddha Mondal
Student
★★★★★
"One of the most wonderful education centres out there. Education is not limited to school syllabus but focuses on skill development."
Vansh Agarwal
Student
★★★★★
"My child Dhairya is really enjoying the Modern Age Coders classes. This is his first online class and he eagerly looks forward to it. I can already see his improvement, and the teachers are very cooperative."
Sonam Oswal
Parent of Dhairya
★★★★★
"Modern Age Coders have wonderful teachers who teach in a clear, easy and practical way. The teacher boosts students' confidence and inspires them to learn without hesitation."
Sonu Goyal
Parent
★★★★★
"I highly recommend this computer coding class! The teachers are incredibly knowledgeable and passionate about coding."
Ritu Kedia
Parent
The first task might be a circle drawn round a house on a map of a playground, a function that reports how much of an area is edge, or a regional measure from work that has been ranking places by their size for years.
The district, its six buurten, one boundary problem, the teaching and the terms.
That question cannot be answered from the published figures, and the numbers that look as though they answer it do not. Six primary establishments in the centre teach 1,989 pupils and the district has 3,130 residents under fifteen; in Prins Alexander it is 28 establishments, 8,108 pupils and 14,585 children. Dividing one by the other gives 0.64 and 0.56, but those divide people attending schools inside a boundary by people living inside it, which are different populations. No catchment data was found, so this page does not pretend to know.
Because a count made inside a boundary throws away whatever sits just outside it, and the smaller the area the more of its residents are near an edge. Modelling the centre as a square of its 413 hectares of land gives sides of about 2.03 kilometres, so only about 26 percent of it lies more than five hundred metres from a border. The same calculation for Prins Alexander at 1,738 hectares gives about 58 percent. That is a fact about shape, not about schools, and it biases any within-boundary measure of what is nearby.
Six buurten: the Cs Kwartier, the Stadsdriehoek, Cool, the Oude Westen, Dijkzigt and the Nieuwe Werk. Together they hold 42,924 residents on 488 hectares, of which 413 are land and 75 water, at a published density of 10,191 people to the square kilometre, which makes this the densest and the smallest of the fourteen gebieden.
3,130 residents are under fifteen, which is 7.3 percent of the district and the lowest share of any page in this series. 18,750 residents, 44 percent, are between twenty-five and forty-five, and 15,290 of the 25,035 households hold one person. It is a district of working-age adults living mostly alone in rented flats, 17,139 of the 23,478 dwellings being rented.
No, it runs in English from the first minute. Four Dutch school words are held back from that rule, groep, havo, vwo and profielkeuze, because a child recognises them from their own timetable and an English rendering would only have to be translated back. No course is offered in a Dutch edition.
The teacher is three and a half hours ahead for part of the year and four and a half for the rest, a difference caused entirely by the Dutch clock changing. In practice that leaves the afternoons after school, the early evenings and the whole of both weekend days. One repeating hour is agreed during the free lesson and then kept.
Because of the arithmetic in the district itself. A group of five to ten learners at one specific level in one specific subject needs a pool, and a district with 3,130 children under fifteen across four square kilometres does not have one, however good its schools are. Groups here are assembled on ability from across the world, which is the only arrangement in which a narrow level actually fills.
No, and none is claimed in the centre, elsewhere in Rotterdam or anywhere in the Netherlands. All teaching is live and online and needs a laptop or desktop, working audio, a current browser and a connection that holds. The telephone number shown on this page is an Indian one and is described as such.
The first lesson is free. After it, group tuition is USD 100 a month for two live lessons a week, normally eight across the month, in a group of five to ten learners. One-to-one tuition is USD 150 a month at the same frequency with a single teacher. Course, format, recurring hour and availability are all confirmed before anything is paid.
Five to ten, matched on level, pace and goal rather than on age, school or address. One-to-one means a single learner with a single teacher. When no group at the right level is running, a family can take one-to-one lessons or wait for the next group to form.
The city page, coding classes in Rotterdam, carries the port throughput argument and the map of all fourteen gebieden, and the other district pages follow this one in the series. North up the line are Delft and The Hague, and everything here sits inside Zuid-Holland. The whole series descends from the Netherlands coding hub, and a learner who wants trained models rather than running programs should begin at AI and machine learning classes in the Netherlands or the AI and Python Academy.