diff --git a/_site/archive.html b/_site/archive.html
index 55d976a..648594c 100644
--- a/_site/archive.html
+++ b/_site/archive.html
@@ -6,7 +6,7 @@
Dimitri Lozeve - Archives
- Background: optimal transport
The essential backbone of the method is the Wasserstein distance, derived from optimal transport theory. Optimal transport is a fascinating and deep subject, so I won’t enter into the details here. For an introduction to the theory and its applications, check out the excellent book from Peyré and Cuturi (2019 ) , (available on ArXiv as well). There are also very nice posts (in French) by Gabriel Peyré on the CNRS maths blog . Many more resources (including slides for presentations) are available at https://optimaltransport.github.io . For a more complete theoretical treatment of the subject, check out Santambrogio (2015 ) , or, if you’re feeling particularly adventurous, Villani (2009 ) .
-For this paper, only a superficial understanding of how the Wasserstein distance works is necessary. Optimal transport is an optimisation technique to lift a distance between points in a given metric space, to a distance between probability distributions over this metric space. The historical example is to move piles of dirt around: you know the distance between any two points, and you have piles of dirt lying around. Now, if you want to move these piles to another configuration (fewer piles, say, or a different repartition of dirt a few metres away), you need to find the most efficient way to move them. The total cost you obtain will define a distance between the two configurations of dirt, and is usually called the earth mover’s distance , which is just an instance of the general Wasserstein metric.
+For this paper, only a superficial understanding of how the Wasserstein distance works is necessary. Optimal transport is an optimisation technique to lift a distance between points in a given metric space, to a distance between probability distributions over this metric space. The historical example is to move piles of dirt around: you know the distance between any two points, and you have piles of dirt lying around⊕ Optimal transport originated with Monge, and then Kantorovich, both of whom had very clear military applications in mind (either in Revolutionary France, or during WWII). A lot of historical examples move cannon balls, or other military equipment, along a front line. earth mover’s distance , which is just an instance of the general Wasserstein metric.
More formally, we start with two sets of points \(x = (x_1, x_2, \ldots,
x_n)\) , and \(y = (y_1, y_2, \ldots, y_n)\) , along with probability distributions \(p \in \Delta^n\) , \(q \in \Delta^m\) over \(x\) and \(y\) (\(\Delta^n\) is the probability simplex of dimension \(n\) , i.e. the set of vectors of size \(n\) summing to 1). We can then define the Wasserstein distance as \[
W_1(p, q) = \min_{P \in \mathbb{R}_+^{n\times m}} \sum_{i,j} C_{i,j} P_{i,j}
@@ -97,12 +99,6 @@ W_1(p, q) = \min_{P \in \mathbb{R}_+^{n\times m}} \sum_{i,j} C_{i,j} P_{i,j}
Yurochkin, Mikhail, Sebastian Claici, Edward Chien, Farzaneh Mirzazadeh, and Justin M Solomon. 2019. “Hierarchical Optimal Transport for Document Representation.” In Advances in Neural Information Processing Systems 32 , 1599–1609. http://papers.nips.cc/paper/8438-hierarchical-optimal-transport-for-document-representation.pdf .
-
]]>
diff --git a/_site/contact.html b/_site/contact.html
index deaf540..a9944ba 100644
--- a/_site/contact.html
+++ b/_site/contact.html
@@ -6,7 +6,7 @@
Dimitri Lozeve - Contact
- Dimitri Lozeve - Curriculum Vitæ
- Dimitri Lozeve - Home
- Dimitri Lozeve - Random matrices from the Ginibre ensemble
- Dimitri Lozeve - Reading notes: Hierarchical Optimal Transport for Document Representation
- Background: optimal transport
The essential backbone of the method is the Wasserstein distance, derived from optimal transport theory. Optimal transport is a fascinating and deep subject, so I won’t enter into the details here. For an introduction to the theory and its applications, check out the excellent book from Peyré and Cuturi (2019 ) , (available on ArXiv as well). There are also very nice posts (in French) by Gabriel Peyré on the CNRS maths blog . Many more resources (including slides for presentations) are available at https://optimaltransport.github.io . For a more complete theoretical treatment of the subject, check out Santambrogio (2015 ) , or, if you’re feeling particularly adventurous, Villani (2009 ) .
-For this paper, only a superficial understanding of how the Wasserstein distance works is necessary. Optimal transport is an optimisation technique to lift a distance between points in a given metric space, to a distance between probability distributions over this metric space. The historical example is to move piles of dirt around: you know the distance between any two points, and you have piles of dirt lying around. Now, if you want to move these piles to another configuration (fewer piles, say, or a different repartition of dirt a few metres away), you need to find the most efficient way to move them. The total cost you obtain will define a distance between the two configurations of dirt, and is usually called the earth mover’s distance , which is just an instance of the general Wasserstein metric.
+For this paper, only a superficial understanding of how the Wasserstein distance works is necessary. Optimal transport is an optimisation technique to lift a distance between points in a given metric space, to a distance between probability distributions over this metric space. The historical example is to move piles of dirt around: you know the distance between any two points, and you have piles of dirt lying around⊕ Optimal transport originated with Monge, and then Kantorovich, both of whom had very clear military applications in mind (either in Revolutionary France, or during WWII). A lot of historical examples move cannon balls, or other military equipment, along a front line. earth mover’s distance , which is just an instance of the general Wasserstein metric.
More formally, we start with two sets of points \(x = (x_1, x_2, \ldots,
x_n)\) , and \(y = (y_1, y_2, \ldots, y_n)\) , along with probability distributions \(p \in \Delta^n\) , \(q \in \Delta^m\) over \(x\) and \(y\) (\(\Delta^n\) is the probability simplex of dimension \(n\) , i.e. the set of vectors of size \(n\) summing to 1). We can then define the Wasserstein distance as \[
W_1(p, q) = \min_{P \in \mathbb{R}_+^{n\times m}} \sum_{i,j} C_{i,j} P_{i,j}
@@ -116,12 +118,6 @@ W_1(p, q) = \min_{P \in \mathbb{R}_+^{n\times m}} \sum_{i,j} C_{i,j} P_{i,j}
Yurochkin, Mikhail, Sebastian Claici, Edward Chien, Farzaneh Mirzazadeh, and Justin M Solomon. 2019. “Hierarchical Optimal Transport for Document Representation.” In Advances in Neural Information Processing Systems 32 , 1599–1609. http://papers.nips.cc/paper/8438-hierarchical-optimal-transport-for-document-representation.pdf .
-
diff --git a/_site/posts/ising-apl.html b/_site/posts/ising-apl.html
index 289a19b..4a73d45 100644
--- a/_site/posts/ising-apl.html
+++ b/_site/posts/ising-apl.html
@@ -6,7 +6,7 @@
Dimitri Lozeve - Ising model simulation in APL
- Dimitri Lozeve - Ising model simulation
- Dimitri Lozeve - Generating and representing L-systems
- Dimitri Lozeve - Peano Axioms
- Dimitri Lozeve - Quick Notes on Reinforcement Learning
- Dimitri Lozeve - Mindsay: Towards Self-Learning Chatbots
- Dimitri Lozeve - Projects
- Dimitri Lozeve - ADS-B data visualization
- Dimitri Lozeve - Civilisation
- Dimitri Lozeve - Community Detection
- Dimitri Lozeve - Ising model simulation
- Dimitri Lozeve - L-systems
- Dimitri Lozeve - Orbit
- Dimitri Lozeve - Satrap
- Dimitri Lozeve - Topological Data Analysis of time-dependent networks
- Dimitri Lozeve - WWII bombings visualization
- Background: optimal transport
The essential backbone of the method is the Wasserstein distance, derived from optimal transport theory. Optimal transport is a fascinating and deep subject, so I won’t enter into the details here. For an introduction to the theory and its applications, check out the excellent book from Peyré and Cuturi (2019 ) , (available on ArXiv as well). There are also very nice posts (in French) by Gabriel Peyré on the CNRS maths blog . Many more resources (including slides for presentations) are available at https://optimaltransport.github.io . For a more complete theoretical treatment of the subject, check out Santambrogio (2015 ) , or, if you’re feeling particularly adventurous, Villani (2009 ) .
-For this paper, only a superficial understanding of how the Wasserstein distance works is necessary. Optimal transport is an optimisation technique to lift a distance between points in a given metric space, to a distance between probability distributions over this metric space. The historical example is to move piles of dirt around: you know the distance between any two points, and you have piles of dirt lying around. Now, if you want to move these piles to another configuration (fewer piles, say, or a different repartition of dirt a few metres away), you need to find the most efficient way to move them. The total cost you obtain will define a distance between the two configurations of dirt, and is usually called the earth mover’s distance , which is just an instance of the general Wasserstein metric.
+For this paper, only a superficial understanding of how the Wasserstein distance works is necessary. Optimal transport is an optimisation technique to lift a distance between points in a given metric space, to a distance between probability distributions over this metric space. The historical example is to move piles of dirt around: you know the distance between any two points, and you have piles of dirt lying around⊕ Optimal transport originated with Monge, and then Kantorovich, both of whom had very clear military applications in mind (either in Revolutionary France, or during WWII). A lot of historical examples move cannon balls, or other military equipment, along a front line. earth mover’s distance , which is just an instance of the general Wasserstein metric.
More formally, we start with two sets of points \(x = (x_1, x_2, \ldots,
x_n)\) , and \(y = (y_1, y_2, \ldots, y_n)\) , along with probability distributions \(p \in \Delta^n\) , \(q \in \Delta^m\) over \(x\) and \(y\) (\(\Delta^n\) is the probability simplex of dimension \(n\) , i.e. the set of vectors of size \(n\) summing to 1). We can then define the Wasserstein distance as \[
W_1(p, q) = \min_{P \in \mathbb{R}_+^{n\times m}} \sum_{i,j} C_{i,j} P_{i,j}
@@ -93,12 +95,6 @@ W_1(p, q) = \min_{P \in \mathbb{R}_+^{n\times m}} \sum_{i,j} C_{i,j} P_{i,j}
Yurochkin, Mikhail, Sebastian Claici, Edward Chien, Farzaneh Mirzazadeh, and Justin M Solomon. 2019. “Hierarchical Optimal Transport for Document Representation.” In Advances in Neural Information Processing Systems 32 , 1599–1609. http://papers.nips.cc/paper/8438-hierarchical-optimal-transport-for-document-representation.pdf .
-
]]>
diff --git a/_site/skills.html b/_site/skills.html
index 2bb74b8..0b45a39 100644
--- a/_site/skills.html
+++ b/_site/skills.html
@@ -6,7 +6,7 @@
Dimitri Lozeve - Skills in Statistics, Data Science and Machine Learning
-
+
+
+
+
\ No newline at end of file
diff --git a/css/et-book/et-book-bold-line-figures/et-book-bold-line-figures.ttf b/css/et-book/et-book-bold-line-figures/et-book-bold-line-figures.ttf
new file mode 100755
index 0000000..9798360
Binary files /dev/null and b/css/et-book/et-book-bold-line-figures/et-book-bold-line-figures.ttf differ
diff --git a/css/et-book/et-book-bold-line-figures/et-book-bold-line-figures.woff b/css/et-book/et-book-bold-line-figures/et-book-bold-line-figures.woff
new file mode 100755
index 0000000..c6685a1
Binary files /dev/null and b/css/et-book/et-book-bold-line-figures/et-book-bold-line-figures.woff differ
diff --git a/css/et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.eot b/css/et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.eot
new file mode 100755
index 0000000..88a6c74
Binary files /dev/null and b/css/et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.eot differ
diff --git a/css/et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.svg b/css/et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.svg
new file mode 100755
index 0000000..881a6bd
--- /dev/null
+++ b/css/et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.svg
@@ -0,0 +1,244 @@
+
+
+
+
+
+
+
\ No newline at end of file
diff --git a/css/et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.ttf b/css/et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.ttf
new file mode 100755
index 0000000..9da91de
Binary files /dev/null and b/css/et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.ttf differ
diff --git a/css/et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.woff b/css/et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.woff
new file mode 100755
index 0000000..b0e5b68
Binary files /dev/null and b/css/et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.woff differ
diff --git a/css/et-book/et-book-roman-line-figures/et-book-roman-line-figures.eot b/css/et-book/et-book-roman-line-figures/et-book-roman-line-figures.eot
new file mode 100755
index 0000000..22fff5b
Binary files /dev/null and b/css/et-book/et-book-roman-line-figures/et-book-roman-line-figures.eot differ
diff --git a/css/et-book/et-book-roman-line-figures/et-book-roman-line-figures.svg b/css/et-book/et-book-roman-line-figures/et-book-roman-line-figures.svg
new file mode 100755
index 0000000..c5013b3
--- /dev/null
+++ b/css/et-book/et-book-roman-line-figures/et-book-roman-line-figures.svg
@@ -0,0 +1,244 @@
+
+
+
+
+
+
+
\ No newline at end of file
diff --git a/css/et-book/et-book-roman-line-figures/et-book-roman-line-figures.ttf b/css/et-book/et-book-roman-line-figures/et-book-roman-line-figures.ttf
new file mode 100755
index 0000000..daceffb
Binary files /dev/null and b/css/et-book/et-book-roman-line-figures/et-book-roman-line-figures.ttf differ
diff --git a/css/et-book/et-book-roman-line-figures/et-book-roman-line-figures.woff b/css/et-book/et-book-roman-line-figures/et-book-roman-line-figures.woff
new file mode 100755
index 0000000..440ae04
Binary files /dev/null and b/css/et-book/et-book-roman-line-figures/et-book-roman-line-figures.woff differ
diff --git a/css/et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.eot b/css/et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.eot
new file mode 100755
index 0000000..4ef9973
Binary files /dev/null and b/css/et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.eot differ
diff --git a/css/et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.svg b/css/et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.svg
new file mode 100755
index 0000000..24cf3ef
--- /dev/null
+++ b/css/et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.svg
@@ -0,0 +1,244 @@
+
+
+
+
+
+
+
\ No newline at end of file
diff --git a/css/et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.ttf b/css/et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.ttf
new file mode 100755
index 0000000..5ae5198
Binary files /dev/null and b/css/et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.ttf differ
diff --git a/css/et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.woff b/css/et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.woff
new file mode 100755
index 0000000..10eb336
Binary files /dev/null and b/css/et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.woff differ
diff --git a/css/et-book/et-book-semi-bold-old-style-figures/et-book-semi-bold-old-style-figures.eot b/css/et-book/et-book-semi-bold-old-style-figures/et-book-semi-bold-old-style-figures.eot
new file mode 100755
index 0000000..bcc6f9d
Binary files /dev/null and b/css/et-book/et-book-semi-bold-old-style-figures/et-book-semi-bold-old-style-figures.eot differ
diff --git a/css/et-book/et-book-semi-bold-old-style-figures/et-book-semi-bold-old-style-figures.svg b/css/et-book/et-book-semi-bold-old-style-figures/et-book-semi-bold-old-style-figures.svg
new file mode 100755
index 0000000..a6c168a
--- /dev/null
+++ b/css/et-book/et-book-semi-bold-old-style-figures/et-book-semi-bold-old-style-figures.svg
@@ -0,0 +1,243 @@
+
+
+
+
+
+
+
\ No newline at end of file
diff --git a/css/et-book/et-book-semi-bold-old-style-figures/et-book-semi-bold-old-style-figures.ttf b/css/et-book/et-book-semi-bold-old-style-figures/et-book-semi-bold-old-style-figures.ttf
new file mode 100755
index 0000000..725af42
Binary files /dev/null and b/css/et-book/et-book-semi-bold-old-style-figures/et-book-semi-bold-old-style-figures.ttf differ
diff --git a/css/et-book/et-book-semi-bold-old-style-figures/et-book-semi-bold-old-style-figures.woff b/css/et-book/et-book-semi-bold-old-style-figures/et-book-semi-bold-old-style-figures.woff
new file mode 100755
index 0000000..5753ca4
Binary files /dev/null and b/css/et-book/et-book-semi-bold-old-style-figures/et-book-semi-bold-old-style-figures.woff differ
diff --git a/css/tufte.css b/css/tufte.css
new file mode 100644
index 0000000..2ad645b
--- /dev/null
+++ b/css/tufte.css
@@ -0,0 +1,469 @@
+@charset "UTF-8";
+
+/* Import ET Book styles
+ adapted from https://github.com/edwardtufte/et-book/blob/gh-pages/et-book.css */
+
+@font-face {
+ font-family: "et-book";
+ src: url("et-book/et-book-roman-line-figures/et-book-roman-line-figures.eot");
+ src: url("et-book/et-book-roman-line-figures/et-book-roman-line-figures.eot?#iefix") format("embedded-opentype"), url("et-book/et-book-roman-line-figures/et-book-roman-line-figures.woff") format("woff"), url("et-book/et-book-roman-line-figures/et-book-roman-line-figures.ttf") format("truetype"), url("et-book/et-book-roman-line-figures/et-book-roman-line-figures.svg#etbookromanosf") format("svg");
+ font-weight: normal;
+ font-style: normal;
+ font-display: swap;
+}
+
+@font-face {
+ font-family: "et-book";
+ src: url("et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.eot");
+ src: url("et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.eot?#iefix") format("embedded-opentype"), url("et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.woff") format("woff"), url("et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.ttf") format("truetype"), url("et-book/et-book-display-italic-old-style-figures/et-book-display-italic-old-style-figures.svg#etbookromanosf") format("svg");
+ font-weight: normal;
+ font-style: italic;
+ font-display: swap;
+}
+
+@font-face {
+ font-family: "et-book";
+ src: url("et-book/et-book-bold-line-figures/et-book-bold-line-figures.eot");
+ src: url("et-book/et-book-bold-line-figures/et-book-bold-line-figures.eot?#iefix") format("embedded-opentype"), url("et-book/et-book-bold-line-figures/et-book-bold-line-figures.woff") format("woff"), url("et-book/et-book-bold-line-figures/et-book-bold-line-figures.ttf") format("truetype"), url("et-book/et-book-bold-line-figures/et-book-bold-line-figures.svg#etbookromanosf") format("svg");
+ font-weight: bold;
+ font-style: normal;
+ font-display: swap;
+}
+
+@font-face {
+ font-family: "et-book-roman-old-style";
+ src: url("et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.eot");
+ src: url("et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.eot?#iefix") format("embedded-opentype"), url("et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.woff") format("woff"), url("et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.ttf") format("truetype"), url("et-book/et-book-roman-old-style-figures/et-book-roman-old-style-figures.svg#etbookromanosf") format("svg");
+ font-weight: normal;
+ font-style: normal;
+ font-display: swap;
+}
+
+/* Tufte CSS styles */
+html {
+ font-size: 15px;
+}
+
+body {
+ width: 87.5%;
+ margin-left: auto;
+ margin-right: auto;
+ padding-left: 12.5%;
+ font-family: et-book, Palatino, "Palatino Linotype", "Palatino LT STD", "Book Antiqua", Georgia, serif;
+ background-color: #fffff8;
+ color: #111;
+ max-width: 1400px;
+ counter-reset: sidenote-counter;
+}
+
+h1 {
+ font-weight: 400;
+ margin-top: 4rem;
+ margin-bottom: 1.5rem;
+ font-size: 3.2rem;
+ line-height: 1;
+}
+
+h2 {
+ font-style: italic;
+ font-weight: 400;
+ margin-top: 2.1rem;
+ margin-bottom: 1.4rem;
+ font-size: 2.2rem;
+ line-height: 1;
+}
+
+h3 {
+ font-style: italic;
+ font-weight: 400;
+ font-size: 1.7rem;
+ margin-top: 2rem;
+ margin-bottom: 1.4rem;
+ line-height: 1;
+}
+
+hr {
+ display: block;
+ height: 1px;
+ width: 55%;
+ border: 0;
+ border-top: 1px solid #ccc;
+ margin: 1em 0;
+ padding: 0;
+}
+
+p.subtitle {
+ font-style: italic;
+ margin-top: 1rem;
+ margin-bottom: 1rem;
+ font-size: 1.8rem;
+ display: block;
+ line-height: 1;
+}
+
+.numeral {
+ font-family: et-book-roman-old-style;
+}
+
+.danger {
+ color: red;
+}
+
+article {
+ padding: 5rem 0rem;
+}
+
+section {
+ padding-top: 1rem;
+ padding-bottom: 1rem;
+}
+
+p,
+ol,
+ul {
+ font-size: 1.4rem;
+ line-height: 2rem;
+}
+
+p {
+ margin-top: 1.4rem;
+ margin-bottom: 1.4rem;
+ padding-right: 0;
+ vertical-align: baseline;
+}
+
+/* Chapter Epigraphs */
+div.epigraph {
+ margin: 5em 0;
+}
+
+div.epigraph > blockquote {
+ margin-top: 3em;
+ margin-bottom: 3em;
+}
+
+div.epigraph > blockquote,
+div.epigraph > blockquote > p {
+ font-style: italic;
+}
+
+div.epigraph > blockquote > footer {
+ font-style: normal;
+}
+
+div.epigraph > blockquote > footer > cite {
+ font-style: italic;
+}
+/* end chapter epigraphs styles */
+
+blockquote {
+ font-size: 1.4rem;
+}
+
+blockquote p {
+ width: 55%;
+ margin-right: 40px;
+}
+
+blockquote footer {
+ width: 55%;
+ font-size: 1.1rem;
+ text-align: right;
+}
+
+section > p,
+section > footer,
+section > table {
+ width: 55%;
+}
+
+/* 50 + 5 == 55, to be the same width as paragraph */
+section > ol,
+section > ul {
+ width: 50%;
+ -webkit-padding-start: 5%;
+}
+
+li:not(:first-child) {
+ margin-top: 0.25rem;
+}
+
+figure {
+ padding: 0;
+ border: 0;
+ font-size: 100%;
+ font: inherit;
+ vertical-align: baseline;
+ max-width: 55%;
+ -webkit-margin-start: 0;
+ -webkit-margin-end: 0;
+ margin: 0 0 3em 0;
+}
+
+figcaption {
+ float: right;
+ clear: right;
+ margin-top: 0;
+ margin-bottom: 0;
+ font-size: 1.1rem;
+ line-height: 1.6;
+ vertical-align: baseline;
+ position: relative;
+ max-width: 40%;
+}
+
+figure.fullwidth figcaption {
+ margin-right: 24%;
+}
+
+/* Links: replicate underline that clears descenders */
+a:link,
+a:visited {
+ color: inherit;
+}
+
+.no-tufte-underline:link {
+ background: unset;
+ text-shadow: unset;
+}
+
+a:link, .tufte-underline, .hover-tufte-underline:hover {
+ text-decoration: none;
+ background: -webkit-linear-gradient(#fffff8, #fffff8), -webkit-linear-gradient(#fffff8, #fffff8), -webkit-linear-gradient(currentColor, currentColor);
+ background: linear-gradient(#fffff8, #fffff8), linear-gradient(#fffff8, #fffff8), linear-gradient(currentColor, currentColor);
+ -webkit-background-size: 0.05em 1px, 0.05em 1px, 1px 1px;
+ -moz-background-size: 0.05em 1px, 0.05em 1px, 1px 1px;
+ background-size: 0.05em 1px, 0.05em 1px, 1px 1px;
+ background-repeat: no-repeat, no-repeat, repeat-x;
+ text-shadow: 0.03em 0 #fffff8, -0.03em 0 #fffff8, 0 0.03em #fffff8, 0 -0.03em #fffff8, 0.06em 0 #fffff8, -0.06em 0 #fffff8, 0.09em 0 #fffff8, -0.09em 0 #fffff8, 0.12em 0 #fffff8, -0.12em 0 #fffff8, 0.15em 0 #fffff8, -0.15em 0 #fffff8;
+ background-position: 0% 93%, 100% 93%, 0% 93%;
+}
+
+@media screen and (-webkit-min-device-pixel-ratio: 0) {
+ a:link, .tufte-underline, .hover-tufte-underline:hover {
+ background-position-y: 87%, 87%, 87%;
+ }
+}
+
+a:link::selection,
+a:link::-moz-selection {
+ text-shadow: 0.03em 0 #b4d5fe, -0.03em 0 #b4d5fe, 0 0.03em #b4d5fe, 0 -0.03em #b4d5fe, 0.06em 0 #b4d5fe, -0.06em 0 #b4d5fe, 0.09em 0 #b4d5fe, -0.09em 0 #b4d5fe, 0.12em 0 #b4d5fe, -0.12em 0 #b4d5fe, 0.15em 0 #b4d5fe, -0.15em 0 #b4d5fe;
+ background: #b4d5fe;
+}
+
+/* Sidenotes, margin notes, figures, captions */
+img {
+ max-width: 100%;
+}
+
+.sidenote,
+.marginnote {
+ float: right;
+ clear: right;
+ margin-right: -60%;
+ width: 50%;
+ margin-top: 0;
+ margin-bottom: 0;
+ font-size: 1.1rem;
+ line-height: 1.3;
+ vertical-align: baseline;
+ position: relative;
+}
+
+.sidenote-number {
+ counter-increment: sidenote-counter;
+}
+
+.sidenote-number:after,
+.sidenote:before {
+ font-family: et-book-roman-old-style;
+ position: relative;
+ vertical-align: baseline;
+}
+
+.sidenote-number:after {
+ content: counter(sidenote-counter);
+ font-size: 1rem;
+ top: -0.5rem;
+ left: 0.1rem;
+}
+
+.sidenote:before {
+ content: counter(sidenote-counter) " ";
+ font-size: 1rem;
+ top: -0.5rem;
+}
+
+blockquote .sidenote,
+blockquote .marginnote {
+ margin-right: -82%;
+ min-width: 59%;
+ text-align: left;
+}
+
+div.fullwidth,
+table.fullwidth {
+ width: 100%;
+}
+
+div.table-wrapper {
+ overflow-x: auto;
+ font-family: "Trebuchet MS", "Gill Sans", "Gill Sans MT", sans-serif;
+}
+
+.sans {
+ font-family: "Gill Sans", "Gill Sans MT", Calibri, sans-serif;
+ letter-spacing: .03em;
+}
+
+code, pre > code {
+ font-family: Consolas, "Liberation Mono", Menlo, Courier, monospace;
+ font-size: 1.0rem;
+ line-height: 1.42;
+ -webkit-text-size-adjust: 100%; /* Prevent adjustments of font size after orientation changes in iOS. See https://github.com/edwardtufte/tufte-css/issues/81#issuecomment-261953409 */
+}
+
+.sans > code {
+ font-size: 1.2rem;
+}
+
+h1 > code,
+h2 > code,
+h3 > code {
+ font-size: 0.80em;
+}
+
+.marginnote > code,
+.sidenote > code {
+ font-size: 1rem;
+}
+
+pre > code {
+ font-size: 0.9rem;
+ width: 52.5%;
+ margin-left: 2.5%;
+ overflow-x: auto;
+ display: block;
+}
+
+pre.fullwidth > code {
+ width: 90%;
+}
+
+.fullwidth {
+ max-width: 90%;
+ clear:both;
+}
+
+span.newthought {
+ font-variant: small-caps;
+ font-size: 1.2em;
+}
+
+input.margin-toggle {
+ display: none;
+}
+
+label.sidenote-number {
+ display: inline;
+}
+
+label.margin-toggle:not(.sidenote-number) {
+ display: none;
+}
+
+.iframe-wrapper {
+ position: relative;
+ padding-bottom: 56.25%; /* 16:9 */
+ padding-top: 25px;
+ height: 0;
+}
+
+.iframe-wrapper iframe {
+ position: absolute;
+ top: 0;
+ left: 0;
+ width: 100%;
+ height: 100%;
+}
+
+@media (max-width: 760px) {
+ body {
+ width: 84%;
+ padding-left: 8%;
+ padding-right: 8%;
+ }
+
+ hr,
+ section > p,
+ section > footer,
+ section > table {
+ width: 100%;
+ }
+
+ pre > code {
+ width: 97%;
+ }
+
+ section > ol {
+ width: 90%;
+ }
+
+ section > ul {
+ width: 90%;
+ }
+
+ figure {
+ max-width: 90%;
+ }
+
+ figcaption,
+ figure.fullwidth figcaption {
+ margin-right: 0%;
+ max-width: none;
+ }
+
+ blockquote {
+ margin-left: 1.5em;
+ margin-right: 0em;
+ }
+
+ blockquote p,
+ blockquote footer {
+ width: 100%;
+ }
+
+ label.margin-toggle:not(.sidenote-number) {
+ display: inline;
+ }
+
+ .sidenote,
+ .marginnote {
+ display: none;
+ }
+
+ .margin-toggle:checked + .sidenote,
+ .margin-toggle:checked + .marginnote {
+ display: block;
+ float: left;
+ left: 1rem;
+ clear: both;
+ width: 95%;
+ margin: 1rem 2.5%;
+ vertical-align: baseline;
+ position: relative;
+ }
+
+ label {
+ cursor: pointer;
+ }
+
+ div.table-wrapper,
+ table {
+ width: 85%;
+ }
+
+ img {
+ width: 100%;
+ }
+}
diff --git a/newblog.cabal b/newblog.cabal
index 9981f7e..a3aceb2 100644
--- a/newblog.cabal
+++ b/newblog.cabal
@@ -9,6 +9,7 @@ executable site
, containers
, pandoc
, pandoc-citeproc
+ , pandoc-sidenote
, hakyll
ghc-options: -threaded
default-language: Haskell2010
diff --git a/posts/hierarchical-optimal-transport-for-document-classification.org b/posts/hierarchical-optimal-transport-for-document-classification.org
index 5fc4e55..1d39f19 100644
--- a/posts/hierarchical-optimal-transport-for-document-classification.org
+++ b/posts/hierarchical-optimal-transport-for-document-classification.org
@@ -63,7 +63,7 @@ the two configurations of dirt, and is usually called the [[https://en.wikipedia
mover's distance]], which is just an instance of the general Wasserstein
metric.
-[fn:historical_ot] Optimal transport originated with Monge, and then
+[fn:historical_ot] {-} Optimal transport originated with Monge, and then
Kantorovich, both of whom had very clear military applications in mind
(either in Revolutionary France, or during WWII). A lot of historical
examples move cannon balls, or other military equipment, along a front
diff --git a/site.hs b/site.hs
index 77b596a..263a62a 100644
--- a/site.hs
+++ b/site.hs
@@ -8,6 +8,7 @@ import Text.CSL.Pandoc (processCites)
import Text.Pandoc
import Text.Pandoc.Options
import Text.Pandoc.Highlighting
+import Text.Pandoc.SideNote
import Hakyll
import Hakyll.Core.Compiler.Internal (compilerProvider, compilerAsk)
@@ -169,7 +170,7 @@ customPandocCompiler =
csl <- load $ fromFilePath "csl/chicago-author-date.csl"
bib <- load $ fromFilePath "bib/bibliography.bib"
writePandocWith writerOptions <$>
- (getResourceBody >>= myReadPandocBiblio readerOptions csl bib)
+ (getResourceBody >>= myReadPandocBiblio readerOptions csl bib >>= traverse (return . usingSideNotes))
type FeedRenderer = FeedConfiguration
-> Context String
diff --git a/stack.yaml b/stack.yaml
index 6ad76a0..ae10b64 100644
--- a/stack.yaml
+++ b/stack.yaml
@@ -37,7 +37,9 @@ packages:
# Dependency packages to be pulled from upstream that are not in the resolver
# using the same syntax as the packages field.
# (e.g., acme-missiles-0.3)
-# extra-deps: []
+extra-deps:
+ - pandoc-sidenote-0.19.0.0@sha256:f5a5fdab1900da7b26ca673d14302b0a27b56024ca811babcacab79ba9614e90,1196
+ - monad-gen-0.3.0.1@sha256:a2569465cfbd468d3350ef25de56b3362580e77537224313aab1210f40804a3b,821
# Override default flag values for local packages and extra-deps
# flags: {}
diff --git a/stack.yaml.lock b/stack.yaml.lock
index 296b4d0..261e0cb 100644
--- a/stack.yaml.lock
+++ b/stack.yaml.lock
@@ -3,7 +3,21 @@
# For more information, please see the documentation at:
# https://docs.haskellstack.org/en/stable/lock_files
-packages: []
+packages:
+- completed:
+ hackage: pandoc-sidenote-0.19.0.0@sha256:f5a5fdab1900da7b26ca673d14302b0a27b56024ca811babcacab79ba9614e90,1196
+ pantry-tree:
+ size: 273
+ sha256: 0d23026f86da13ca57aaa0041bfe719cf7acbeddbadbbe1bb8e265316dcfce49
+ original:
+ hackage: pandoc-sidenote-0.19.0.0@sha256:f5a5fdab1900da7b26ca673d14302b0a27b56024ca811babcacab79ba9614e90,1196
+- completed:
+ hackage: monad-gen-0.3.0.1@sha256:a2569465cfbd468d3350ef25de56b3362580e77537224313aab1210f40804a3b,821
+ pantry-tree:
+ size: 288
+ sha256: a825dc8dcc3ee27dbbc1b71f6853703790ed320249b8f8510bb36d726664e795
+ original:
+ hackage: monad-gen-0.3.0.1@sha256:a2569465cfbd468d3350ef25de56b3362580e77537224313aab1210f40804a3b,821
snapshots:
- completed:
size: 506303
diff --git a/templates/default.html b/templates/default.html
index 33f3717..8c7439f 100644
--- a/templates/default.html
+++ b/templates/default.html
@@ -6,7 +6,7 @@
Dimitri Lozeve - $title$
-