Jon Kågström

Using floating point arithmetic, in particular multiplication and division with many parameters, will sooner or later lead to either underflow or overflow (not enough bits to represent the number). This article shows a trick that in some cases can be very useful to avoid precision loss, the sum of logs. (more…)

Can you see why the code at the bottom is working as expected for integers (integral numbers) but not doubles (or other floating point numbers for that sake)? If not, don’t worry, you are not alone. (more…)

When programming C++ it is a very common need to switch between the .h and .cpp file. There are no built in way to do this in Visual Studio, however it is possible to accomplish! You can either use an expensive helper program, such as Visual Assist or you can do it freely by writing a macro. (more…)