designsense
46 of 46 options
Brush a chart or drag over images; brush again for OR. Click a point, line or image to inspect.
Navigate
Save & compare
Setup & share
Parallel coordinates
Colour
199 40k
Drag along an axis to filter · drag a title to reorder, double-click it to flip its axis
40k35k30k25k20k15k10k5kprice ↓Lower is better, so this axis runs from high at the foot to low at the top.1950196019701980199020002010year ↑012345678cylinders500k400k300k200k100k0odometer ↓Lower is better, so this axis runs from high at the foot to low at the top.
Scatterplot
X
Y
price ↓10k20k30k40kyear ↑1950196019701980199020002010
price
year
cylinders
odometer
Decide · Goals and your own criteria
Set which direction is better for each output. In this example, a lower price and a later year are better. Arrows on the axes then show the better direction, and DesignSense can find the best trade-offs. Further down you can add criteria that are not in the data. A custom dimension holds values you enter by hand, such as a score for looks. A weighted dimension combines several columns into one score.
Dimensions and Goals
Goals
price↓ from the sample
year↑ from the sample
cylinders
odometer↓ from the sample
Custom and Weighted Dimensions

Judge options on criteria of your own with custom dimensions, then weigh several together into a weighted dimension for multicriteria decision analysis. Enter values beside the star rating or by dragging a knob on the parallel coordinates.

No custom or weighted dimensions yet.

These follow the design space construction framework of Haymaker et al. (2018), in which options are judged against objectives and the objectives are weighed by preference into an overall value: custom dimensions are objectives you judge yourself, and weighted dimensions are that value. See “Design space construction: a framework to support collaborative, parametric decision making,” ITcon 23(8): 157–178.