Skip to content

Commit

Permalink
Added def of set size in equiv chapter
Browse files Browse the repository at this point in the history
  • Loading branch information
whitead committed Jul 6, 2022
1 parent def94dc commit 40555c1
Show file tree
Hide file tree
Showing 2 changed files with 3 additions and 3 deletions.
2 changes: 1 addition & 1 deletion _static/custom.js
Expand Up @@ -87,6 +87,6 @@ function autoplayVideos() {
})
}

window.addEventListener('load', autoplayVideos)
window.addEventListener('load', addImgAnchors)
window.addEventListener('load', autoplayVideos)
window.addEventListener('load', addGithubLink)
4 changes: 2 additions & 2 deletions dl/Equivariant.ipynb
Expand Up @@ -404,7 +404,7 @@
"\n",
"## Converting between Space and Group\n",
"\n",
"Let's see how we can convert between functions on the space $\\mathcal{X}$ and functions on the group $G$. $|G| \\geq |\\mathcal{X}|$ (because the space is homogeneous) so it is rare that we can uniquely replace each point in space with a group in $G$. Instead, we will construct a partitioning of $G$ into $|\\mathcal{X}|$ sets called a quotient space $G / H$ such that $|G / H| = |\\mathcal{X}|$. It turns out, there is a well-studied approach to arranging elements in a group called **cosets**. Constructing cosets is a two-step process. First we define a subgroup $H$. A **subgroup** means it is itself a group; it has identities and inverses. We cannot accidentally leave $H$, $h_1\\cdot{} h_2 \\in H$. For example, translation transformations are a subgroup because you cannot accidentally create a rotation when combining two translations. \n",
"Let's see how we can convert between functions on the space $\\mathcal{X}$ and functions on the group $G$. $|G| \\geq |\\mathcal{X}|$ ($|G|$ is number of elements) because the space is homogeneous, so it is rare that we can uniquely replace each point in space with a group in $G$. Instead, we will construct a partitioning of $G$ into $|\\mathcal{X}|$ sets called a quotient space $G / H$ such that $|G / H| = |\\mathcal{X}|$. It turns out, there is a well-studied approach to arranging elements in a group called **cosets**. Constructing cosets is a two-step process. First we define a subgroup $H$. A **subgroup** means it is itself a group; it has identities and inverses. We cannot accidentally leave $H$, $h_1\\cdot{} h_2 \\in H$. For example, translation transformations are a subgroup because you cannot accidentally create a rotation when combining two translations. \n",
"\n",
"```{margin}\n",
"This process of constructing cosets and then using that to lift our function is closely related to the process of finding an induced representation on $G$ via a representation on $H$.\n",
Expand Down Expand Up @@ -2039,7 +2039,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.8.12"
"version": "3.8.5"
}
},
"nbformat": 4,
Expand Down

0 comments on commit 40555c1

Please sign in to comment.