All Visual Studio blogs in one place

AddThis Social Bookmark Button
The first time I ran my histogram visualizer I asked for a Cauchy distribution with a minimum of -10 and a maximum of 10, and of course I got a graph that looks much like the one from my article of last week:   Looks perfectly reasonable; I guess my program is correct right out of the gate, because I am that awesome! Then I went to make a graph of a uniform distribution with a minimum of zero and a maximum of one, but I forgot to update the actual query; it still gave me a
Cauchy distribution. Here's that same Cauchy distribution this time graphed only from 0 to 1. Oh, the pain: Which is obviously neither uniform nor Cauchy. Equally obvious: I am not sufficiently awesome to write a twenty-line program without a trivial floating point bug the first time. The bug, which is very subtle in the first graph, was now obvious: the calculation to determine what the count is for the leftmost bucket is wrong. Why? Because converting a double to an integer simply discards the fractional part, effectively truncating towards zero, and "towards zero" is not downwards if...(Read whole news on source site)



Enter your email address:

Delivered by FeedBurner

Increase your website traffic with


Anti-spam: How many eyes has a typical person?

Follow us on FB

Home : Blog List : Fabulous Adventures In Coding : The Solution To The Simple Puzzle