From 64b876ef6c1a2d7aa4136bed0ff3ceac92f81b0a Mon Sep 17 00:00:00 2001
From: Dan Foreman-Mackey <foreman.mackey@gmail.com>
Date: Mon, 28 Mar 2022 05:16:17 -0400
Subject: [PATCH] update changelog (#76)

---
 docs/news.rst   | 31 +++++++++++++++++++++++++++++++
 news/57.bugfix  |  2 --
 news/58.doc     |  2 --
 news/58.feature |  2 --
 news/59.feature |  1 -
 news/60.feature |  1 -
 news/67.feature |  2 --
 news/70.doc     |  3 ---
 8 files changed, 31 insertions(+), 13 deletions(-)
 delete mode 100644 news/57.bugfix
 delete mode 100644 news/58.doc
 delete mode 100644 news/58.feature
 delete mode 100644 news/59.feature
 delete mode 100644 news/60.feature
 delete mode 100644 news/67.feature
 delete mode 100644 news/70.doc

diff --git a/docs/news.rst b/docs/news.rst
index 85df246c..a98c7258 100644
--- a/docs/news.rst
+++ b/docs/news.rst
@@ -5,6 +5,37 @@ Release Notes
 
 .. towncrier release notes start
 
+tinygp 0.2.1 (2022-03-28)
+-------------------------
+
+Features
+~~~~~~~~
+
+- Renamed elements of quasiseparable kernels, and added support for modeling
+  derivative observations with these kernels. (`#58 <https://github.com/dfm/tinygp/issues/58>`_)
+- Added more flexible noise models: diagonal, banded, or dense. (`#59 <https://github.com/dfm/tinygp/issues/59>`_)
+- Added :class:`tinygp.kernels.quasisep.CARMA` kernel to implement CARMA models. (`#60 <https://github.com/dfm/tinygp/issues/60>`_)
+- Added a minimal solver based on Kalman filtering to use as a baseline for
+  checking the performance of the :class:`tinygp.solvers.QuasisepSolver`. (`#67 <https://github.com/dfm/tinygp/issues/67>`_)
+
+
+Bugfixes
+~~~~~~~~
+
+- Fixed exception when conditioning with quasiseparable solver, since quasisep
+  kernels are not hashable. (`#57 <https://github.com/dfm/tinygp/issues/57>`_)
+
+
+Improved Documentation
+~~~~~~~~~~~~~~~~~~~~~~
+
+- Added a new tutorial describing how to model multiband and derivative
+  observations using quasiseparable kernels. (`#58 <https://github.com/dfm/tinygp/issues/58>`_)
+- Add more details to Deep Kernel learning tutorial,
+  showing comparison with Matern-3/2 kernel
+  and the transformed features. (`#70 <https://github.com/dfm/tinygp/issues/70>`_)
+
+
 tinygp 0.2.0 (2022-03-03)
 -------------------------
 
diff --git a/news/57.bugfix b/news/57.bugfix
deleted file mode 100644
index 1f81f620..00000000
--- a/news/57.bugfix
+++ /dev/null
@@ -1,2 +0,0 @@
-Fixed exception when conditioning with quasiseparable solver, since quasisep
-kernels are not hashable.
diff --git a/news/58.doc b/news/58.doc
deleted file mode 100644
index cd78e8ec..00000000
--- a/news/58.doc
+++ /dev/null
@@ -1,2 +0,0 @@
-Added a new tutorial describing how to model multiband and derivative
-observations using quasiseparable kernels.
diff --git a/news/58.feature b/news/58.feature
deleted file mode 100644
index 0a33c052..00000000
--- a/news/58.feature
+++ /dev/null
@@ -1,2 +0,0 @@
-Renamed elements of quasiseparable kernels, and added support for modeling
-derivative observations with these kernels.
diff --git a/news/59.feature b/news/59.feature
deleted file mode 100644
index e2567079..00000000
--- a/news/59.feature
+++ /dev/null
@@ -1 +0,0 @@
-Added more flexible noise models: diagonal, banded, or dense.
diff --git a/news/60.feature b/news/60.feature
deleted file mode 100644
index 5cff1a73..00000000
--- a/news/60.feature
+++ /dev/null
@@ -1 +0,0 @@
-Added :class:`tinygp.kernels.quasisep.CARMA` kernel to implement CARMA models.
diff --git a/news/67.feature b/news/67.feature
deleted file mode 100644
index 86a84ca6..00000000
--- a/news/67.feature
+++ /dev/null
@@ -1,2 +0,0 @@
-Added a minimal solver based on Kalman filtering to use as a baseline for
-checking the performance of the :class:`tinygp.solvers.QuasisepSolver`.
diff --git a/news/70.doc b/news/70.doc
deleted file mode 100644
index 163fdd62..00000000
--- a/news/70.doc
+++ /dev/null
@@ -1,3 +0,0 @@
-Add more details to Deep Kernel learning tutorial,
-showing comparison with Matern-3/2 kernel
-and the transformed features.