When you want tailor them to your own calculations, just replace any references to Quantity with your measure of choice and make sure each of your table calculations are computed using it. Now for the calculations coming, we are using Superstore data to generate box plots of Quantity by State to create the view below. This method uses table calculations that requires index formulae computed using the measure itself so before you start, untick the 'Aggregate Measures' in the analysis menu. So How Do We Calculate This?First of all, the data being plotted needs to be at the most granular level. The WhiskersThe whiskers then stretch out to the furthest data points within 1.5 interquartile ranges of the hinges, where the interquartile range is the distance between the two hinges. Hinges are then the medians of their respective halves of the data. This is illustrated below where the number sets of 1-5, 1-7, and 1-9 each duplicate their median value. However if there are an odd number of data points then the median value is duplicated and used by both halves. If there are an even number of data points then the data is split straight down the middle. Tukey hinges are the midway points in the first and second halves of data. Tukey HingesThe hinges Tableau uses are Tukey inclusionary hinges, so named after John Tukey the person who first created box plots. The hinges however are harder to work out because they are near the 25th and 75th percentiles but not there exactly and the distance away from these values depends on the size of the data set. The median is a standard statistical measure and is included in Tableau's library of formulae, it's the 50th percentile value, the middle number in a data set. Click a dimension, hold ctrl & click a measure, click the 'Show Me' tab and select the box plot function and there you go, you have a box plot! The Constituent PartsBox plots are made of five key components: the median, the upper and lower hinges, and the upper and lower whiskers. In fact the simplest box plot in Tableau takes only 4 clicks. Box plots are great for displaying distribution and in Tableau they're incredibly easy to make.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |