Irregular algorithms are input based computationally heavy algorithms. The input to the algorithm plays a very high impact on how the algorithm behaves and time taken to compute the result. For my research, we have chosen Sparse Fast Fourier Transform (sFFT), and parallelized it for GPUs and multicore systems.

A single bigimg can be specified in the front matter by the following YAML:

bigimg: [{src: "/img/triangle.jpg", desc: "Triangle"}]

Multiple bigimgs can be specified in the front matter by the following YAML:

bigimg: [{src: "/img/triangle.jpg", desc: "Triangle"},
         {src: "/img/sphere.jpg", desc: "Sphere"},
         {src: "/img/hexagon.jpg", desc: "Hexagon"}]

Also note that the description field is optional, and images could instead be specified by:

bigimg: [{src: "/img/triangle.jpg"},
         {src: "/img/sphere.jpg"},
         {src: "/img/hexagon.jpg"}]

The above YAML array of hashes were written in “flow” style. However when generating a new page or post with hugo new post/mypost.md, hugo may interpret the archetype for bigimg in the default YAML style. Defining multiple bigimg’s complete with descriptions in this style would be specified by:

bigimg:
- {src: "/img/triangle.jpg", desc: "Triangle"}
- {src: "/img/sphere.jpg", desc: "Sphere"}
- {src: "/img/hexagon.jpg", desc: "Hexagon"}

Additional information can be found in this YAML tutorial.