Skip to content

Commit

Permalink
Built the .tex.md files
Browse files Browse the repository at this point in the history
  • Loading branch information
potmo committed Feb 7, 2022
1 parent 29a3b28 commit 742b753
Showing 1 changed file with 1 addition and 1 deletion.
2 changes: 1 addition & 1 deletion readme.md
Original file line number Diff line number Diff line change
Expand Up @@ -347,7 +347,7 @@ Even if the De Bruijn sequence is efficient at packing information in a sequence



The radial ordering of the sequences (rings) are generally arbitrary. The sequences closer to the middle cover a smaller area since the circumference increases with radius. Since the mirrors might not be totally exact due to the fabrication process it can be a good idea to have more information density in the periphery of the disc where the area are bigger. We could for example sort the sequences in a way such as we put the single color sequences end up close to the middle and the sequences with multiple colors end up close to the periphery. An over engineering method (that I've opted to go with) is to sort the sequences on their Shannon entropy. Shannon entropy of <img src="svgs/cbfb1b2a33b28eab8a3e59464768e810.svg?invert_in_darkmode" align=middle width=14.908688849999992pt height=22.465723500000017pt/> is defined as: <p align="center"><img src="svgs/88f1c8fd78906178630bcb2103849b3b.svg?invert_in_darkmode" align=middle width=196.01397089999998pt height=44.89738935pt/></p>
The radial ordering of the sequences (rings) are generally arbitrary. The sequences closer to the middle cover a smaller area since the circumference increases with radius. Since the mirrors might not be totally exact due to the fabrication process it can be a good idea to have more information density in the periphery of the disc where the area are bigger. We could for example sort the sequences in a way such as we put the single color sequences end up close to the middle and the sequences with multiple colors end up close to the periphery. An over engineering method (that I've opted to go with) is to sort the sequences on their Shannon entropy. Shannon entropy of <img src="svgs/cbfb1b2a33b28eab8a3e59464768e810.svg?invert_in_darkmode" align=middle width=14.908688849999992pt height=22.465723500000017pt/> is defined as: <p align="center"><img src="svgs/c5ed727db95c51d1dd9f700fb05452df.svg?invert_in_darkmode" align=middle width=204.7659075pt height=44.89738935pt/></p>
Where <img src="svgs/9c5ffbbaadf29257b87ab1639f470396.svg?invert_in_darkmode" align=middle width=62.60089769999999pt height=14.15524440000002pt/> is the possible outcomes which occur with probability <img src="svgs/b63ad1ad63f0e4b88dc75dbcd4c1569a.svg?invert_in_darkmode" align=middle width=111.36816404999999pt height=24.65753399999998pt/>.
We don't really have to dwell too long on the nitty gritty but can conclude we can compute a number for each sequence that will be lower if not much happens in it and high if there is a lot happening. We can then use that to sort the radial rings. In our example here single color sequences ends up in the middle.

Expand Down

0 comments on commit 742b753

Please sign in to comment.