Computations have always been an essential tool in the natural sciences, and the digital computer was originally developed for the purpose of scientific calculations. At most universities, the first computers were therefore obtained for specific scientific calculations.
Computing essential in research — absent in elementary education
The use of computers has spread in sciences, and it has become increasingly common to require that graduate students learn to programme and otherwise utilise computers effectively, often without them having had any formal training previously. Even so, few universities include structured training in programming and advanced use of computers as part of their undergraduate science curriculum.
This was also the situation at the University of Oslo in the late 1990's. There were several groups and individuals whose research depended heavily on simulation and computation, but there was no coordinated computational training, particularly not on undergraduate level.
Introducing computations at an early stage
Due to this, physicists Morten Hjorth-Jensen and Arnt Inge Vistnes decided to try and introduce computations in the elementary mechanics course. Their two main reasons for doing this was that it would expose the students to computations at an early stage, which many of them would need in their graduate studies anyway. Also, and perhaps more importantly so, it would make it possible to apply the physics to real-life problems that could not be tackled with the traditional pencil-and-paper solution methods that the students learnt in their maths classes. Since the average physics student had no background in programming, this meant that the necessary technical skills would have to be learnt in the mechanics course. This course therefore included basic training in Maple.
At the Department of Informatics, there was a large group working in
diverse areas under the general heading ‘Mathematical Modelling’. Students here were taught programming early on, but an increasing proportion of the students included little or no mathematics in their course portfolio. The general feeling was that this was due to the fact that the basic maths courses had no computational component and therefore seemed irrelevant to the students at the Department of Informatics. The Mathematical Modelling group therefore approached the Department of Mathematics with their concern.
It turned out that the mathematicians were already working on adapting their elementary courses more to the different user groups. The result was that in the Autumn of 2000, the first-semester Calculus course was taught in three different versions: MAT100A (the theoretical version), MAT100B (the computational version) and MAT100C (the physics version). The A and C versions were taught by mathematicians, while a teacher from the Mathematical Modelling group taught MAT100B.
The teaching in MAT100B was based on that the students simultaneously attended the basic programming course in Java, and the curriculum included basic programming and analysis of elementary numerical methods. MAT100C (the physics version) included simple Maple exercises. From the start, MAT100B was the most popular of the three alternatives. The other two basic maths courses were also divided into two streams, a theoretical stream and a computational stream, with teachers from the Mathematical Modelling group and the Department of Mathematics sharing the teaching.
New reform in 2003, also CoE - more collaboration
- No specific Physics stream, computational stream instead - learn viatraditional programming course rather than Maple/Matlab
- FAM and MIT together apply for support for introducing unified
- computational perspective from Flexible Learning - granted!
Unified revision for whole faculty in the strategic plan
- Computations in maths courses + programming, physics courses build on this
- Need for elementary programming course with a mathematical perspective, introduced in 2007
- More coordinated computational content in geophysics and astrophysics, also from 2007