Merge pull request #9 from rosalindfranklininstitute/christmastree2023

Christmastree2023

2022/README.md

@@ -0,0 +1,24 @@
+# christmastree2022
+To collate the work that we have done of the Christmas tree designs 2022's Christmas tree decorating competition
+
+# Intro
+
+This year the AI team will submit two Christmas Trees into the competition, the code, issues and docs surrounding the Christmas tree will be gathered here.
+
+## Tree 1: Born-in-a-Stable Diffusion
+
+This tree will have:
+* 3D printed/modelling clay tree ornaments of our Project Animals doing Christmassy things
+* The images to 3D print will be generated via Stable Diffusion and mapped to 3D using this project: https://dreamfusion3d.github.io/
+* We will have the `christmasdream.rfi.ac.uk` website running where users can see the Stable diffusion model outcomes and maybe create their own?
+* A tablet will be next to the tree to display this
+* You can use the star with the QR code to get to `christmasdream.rfi.ac.uk`
+
+## Tree 2: Xmas-tallography
+This tree will have:
+* programmable lights running sinogram patterns
+* Ornaments that are made by laser cutting/Macrame/Crochet/paper-cutting FFT's of christmas images i.e. snowflakes into glass or acrylic
+* The FFT's will be generated from code published in this repo. 
+
+
+

2023/web-deployment/Dockerfile

@@ -0,0 +1,9 @@
+FROM squidfunk/mkdocs-material:9.0.0b3 as build
+
+COPY mkdocs.yml /docs/mkdocs.yml
+COPY docs /docs/docs
+RUN mkdocs build
+
+FROM nginx:stable-alpine
+
+COPY --from=build /docs/site/ /usr/share/nginx/html

2023/web-deployment/docs/assets/stylesheets/extra.css

@@ -0,0 +1,23 @@
+@keyframes snow {
+	0% {
+		fill-opacity: 1;
+	}
+	100% {
+		fill-opacity: 0;
+		transform: translateY(70%);
+	}
+}
+
+body {
+	overflow-x: hidden;
+	color: #080000;
+	margin: 0 auto;
+	background-color: rgba(0, 0, 0, 0.3);
+	background-image: url('santas-workshop.jpg');
+	background-repeat: no-repeat;
+	background-size: cover;
+	background-position: center;
+	background-blend-mode: soft-light;
+	min-height: 100vh;
+}
+

2023/web-deployment/docs/decision-tree.md

@@ -0,0 +1 @@
+A decision tree is a statistical method for making decisions. The tree represents decisions and their consequences as you move down the tree you get to the final decision. You may recognise these methods from magazine quizzes, we have build a decision tree here to decide if you are naughty or nice! Have fun following the questions!

2023/web-deployment/docs/index.md

@@ -0,0 +1,15 @@
+---
+title: Santa's Workshop 2.0 - AI's Christmas Dream 2023 
+ᴴₒᴴₒᴴₒ: true
+---
+
+Earlier this year the RSEs (Research Software Elves) in Santa's workshop were having a dilemma. The old model they used to decide if a child was "Naughty" or "Nice" had been running in production for 10 years and was no longer suitable! They reached out to their collaborators in the AI theme at the Rosalind Franklin Institute to design them a new model to make sure that they could find all the good children and process their letters to Santa in time for Christmas...
+
+The AI Team put their heads together, and came up with a Neural Network to replace the out of date decision tree that the elves had previously been using. The elves asked for the new model to be sent to them via a magic Christmas tree that will deliver the new model on the 1st December.
+
+Our christmas tree this year has two faces:
+
+1. The current Decision Tree used by the elves, play along ... are you Naughty or Nice?
+2. A representation of the new neural network we are sending to replace it. 
+
+Enjoy our entry this year... And Merry Christmas from all at the AI Theme!

2023/web-deployment/docs/neural-network.md

@@ -0,0 +1,7 @@
+A neural network is a type of algorithm that is based on layers of connected nodes. It can makes decisions by evaluating weights at each of the nodes and then passes the result on to the node in the next layer. Each node has a minimisation function that it uses to determine the weights. A network is trained on a sample of data and then is used to make predictions on new data. We know which members of the AI theme won't be getting presents this year, will you?!
+
+Our Christmas tree represents a fully connected neural network (FCNN) where all nodes are connected to all nodes on the precedeing and next layer. The network is a Naughty / Nice classifier to determine if you will get any presents. The nodes are represent by our crafted decorations and the network is represented by tinsel. Our input is you, represented by a mirror! And our output is coal and presents depending on how you have been classified!
+
+![image](./assets/images/christmas_FCNN.png)
+
+A christmas tree showing the representation of a neural network. Input/output shown in blue circles, nodes in layer 1 represented by stars and nodes in layer 2 represneted by red oblongs.

2023/web-deployment/mkdocs.yml

@@ -0,0 +1,16 @@
+site_name: Christmas Dream
+theme:
+  name: material
+  palette:
+    scheme: default
+    primary: indigo
+    accent: teal
+
+  logo: assets/images/logo.png
+  features:
+    - navigation.instant
+    - navigation.tracking
+    - navigation.sections
+extra_css:
+  - assets/stylesheets/extra.css
+

README.md

@@ -1,24 +1,3 @@
-# christmastree2022
-To collate the work that we have done of the Christmas tree designs 2022's Christmas tree decorating competition
+# A collection of ideas and scripts used in the Rosalind Franklin Institute's Christmas tree competion!
 
-# Intro
-
-This year the AI team will submit two Christmas Trees into the competition, the code, issues and docs surrounding the Christmas tree will be gathered here.
-
-## Tree 1: Born-in-a-Stable Diffusion
-
-This tree will have:
-* 3D printed/modelling clay tree ornaments of our Project Animals doing Christmassy things
-* The images to 3D print will be generated via Stable Diffusion and mapped to 3D using this project: https://dreamfusion3d.github.io/
-* We will have the `christmasdream.rfi.ac.uk` website running where users can see the Stable diffusion model outcomes and maybe create their own?
-* A tablet will be next to the tree to display this
-* You can use the star with the QR code to get to `christmasdream.rfi.ac.uk`
-
-## Tree 2: Xmas-tallography
-This tree will have:
-* programmable lights running sinogram patterns
-* Ornaments that are made by laser cutting/Macrame/Crochet/paper-cutting FFT's of christmas images i.e. snowflakes into glass or acrylic
-* The FFT's will be generated from code published in this repo. 
-
-
-
+See each year in a folder