Browse Prior Art Database

Assigning plot colours automatically based on name

IP.com Disclosure Number: IPCOM000132492D
Original Publication Date: 2005-Dec-19
Included in the Prior Art Database: 2005-Dec-19
Document File: 2 page(s) / 94K

Publishing Venue

IBM

Abstract

Disclosed is an improvement to data visualisation methods which makes data representation more meaningful. Most plotting methods assign colours to data-sets automatically based on the relative position of the data-set. The approach described in this disclosure is to assign a colour based on the name of the data-set. This ties the colour to the semantics of the data and results in more meaningful and consistent visualisations.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Assigning plot colours automatically based on name

Disclosed is an improvement to data visualisation methods, such as software programs, which makes the data representation more meaningful. Most plotting programs assign colours automatically based on the relative position of the data-set. The position is matched against a list of pre-defined colours. For example, the first line in a plot will always be magenta, the second green, the third blue, and so on. The plots could be bar charts, pie charts, line-plots, or other data representations. The programs could be spreadsheets, scientific plotting programs, or other data manipulation software.

    Problems arise if plots are generated for sets and sub-sets, or for intersecting data sets. For example, when set {x,a,y,b,z,c} is plotted and set {x,y,z} is plotted separately, the y-line might be blue in the first graph and green in the second graph. If the set {w,x,y} is plotted, y will again be blue. The lack of consistency makes the data visualisation much less effective. Elaborating on the example, {y} might be a column in a spreadsheet which represents sales figures for Yibble Region, and if the Yibble region segment of the pie chart is green when compared to the other Southern Regions (x,y,z), but blue when the Northern Region sales figures are included in the pie chart, viewers might be confused.

    Some programs allow colours to be over-ridden, but usually only by defining a mapping which hard-codes relationships between data sets and colours. For example, y could be permanently associated with the colour maroon. This kind of hard-coding is extra work for the user, and it is is not useful when a new dataset is added until the hard-coding is updated. It is certainly not a general solution to the problem of assigning colours to data-sets in a way which is reproducible when the number or ord...