Update Changelog and fix formatting in some docstrings.

CHANGELOG.md

@@ -5,6 +5,72 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.3.1] - 2021-09-29
+### Added
+- Error raise if geomagnetic field is invalid in class `EKF`.
+- New method `Omega` in class `AQUA` to simplify the product between angular rate and quaternion.
+- New method `rotate_by` in class `QuaternionArray`.
+- More Acronyms in page `Nomeclature` of documentation.
+- Individual pages for each method and attribute of classes `Quaternion` and `DCM` into documentation.
+- Merge pull request of basic automation tests.
+
+### Changed
+- Fix undefined matrix `R` in method `update` of class `EKF`.
+- Fix shape of converted points in function `ned2enu`.
+- Fix parameters in function `rec2geo` of submodule `frames`.
+- Fix method `from_quaternion` of class `DCM`.
+- Fix Munich height (in km) in global constants.
+- Improve detection of empty arrays when building `TRIAD` object.
+- Improve description of estimator `AQUA` in its docstring.
+- Improve imports in submodules `frames` and `dcm`.
+- Improve style and descriptions in docstrings of functions in submodule `orientation`.
+- Method `init_q` is now synonym of the more convenient method `estimate` in class `AQUA`.
+- Parameter `as_quaternion` in method `estimate` of class `TRIAD` renamed to `representation`, and its value is now of type `str`.
+- Versioning is defined using f-strings.
+
+### Removed
+- Redundant normalization of magnetic measurement vector in class `FQA`.
+
+## [0.3.0] - 2021-02-03
+### Added
+- World Magnetic Model as class `WMM`.
+- World Geodetic System as class `WGS`.
+- Geodetic and planetary constants.
+- New class `DCM` to represent Direction Cosine Matrices.
+- Attitude Estimator `EKF`.
+- Attitude Estimator `TRIAD`.
+- Attitude Estimator `Davenport`.
+- Attitude Estimator `QUEST`.
+- Attitude Estimator `SAAM`.
+- Attitude Estimator `OLEQ`.
+- Attitude Estimator `ROLEQ`.
+- Implementation of modes `newton`, `eig` and `symbolic` in estimator `FLAE`.
+- New function `ecompass` to estimate Orientation from an observation of an accelerometer and a magnetometer.
+- New method `row_reduction` in estimator `FLAE`.
+- New methods `to_angles`, `is_pure`, `is_real`, `is_versor`, and `is_identity` in class `QuaternionArray`.
+- New properties `logarithm` and `log` in class `Quaternion`.
+- New parameter `versor` defaulting to `True` in class `Quaternion` to initialize quaternions as versors.
+- New frame transformations `ned2enu`, `enu2ned`, `ecef2enu`, `enu2ecef`, `ll2ecef`, and `ecef2llf`.
+- Requirements file.
+- Manifest file.
+- Documentation with Sphinx.
+- Type hints.
+
+### Changed
+- Fix `AQUA`.
+- Fix missing imports in several submodules.
+- Fix versioning that prevented install from a fresh setup.
+- `Tilt` can perform vectorized estimations over 2D arrays.
+- `QuaternionArray` is now subclassed from a `numpy.array`.
+- Simplify implementation of class `Complementary`.
+- Create copies of quaternions and normalize them when given in `metrics` functions.
+- Complete attitude estimator `FAMC`.
+- Complete docstrings of all functions, methods and classes.
+- Improve examples and information in `README`.
+
+### Removed
+- Submodules `plot` and `io` to remove dependencies on `scipy` and `matplotlib`.
+
 ## [0.2.2-dev1] - 2020-02-06
 ### Added
 - Support for other characters as separators in function `load` of submodule `io`.
@@ -31,9 +97,6 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Class `QuaternionArray` to handle several quaternions at once.
 - Add methods `log()`, `exp()`, `inv()`, `to_array()` and `to_list()` to class `Quaternion`.
 
-### Changed
-- Fix and/or improve 
-
 ## [0.2.0] - 2019-12-27
 ### Added
 - Class `Quaternion` to handle quaternion operations.

README.md

@@ -185,7 +185,7 @@ array([ 0.19537239,  0.17826049, -0.87872408, -0.39736232])
 
 One of the biggest improvements in this version is the addition of many new attitude estimation algorithms.
 
-All estimators are refactored to be consistent to the original articles describing them. They have in-code references to the original equations, so that you can follow the original articles along with the code.
+All estimators are refactored to be consistent with the corresponding articles describing them. They have in-code references to the equations, so that you can follow the original articles along with the code.
 
 Implemented attitude estimators are:
 
@@ -209,7 +209,7 @@ Implemented attitude estimators are:
 | Tilt          | NO        | YES           | Optional     |
 | TRIAD         | NO        | YES           | YES          |
 
-More Estimators are still a *Work In Progress*, or *planned* to be added in the future.
+More Estimators are still a _Work In Progress_, or _planned_ to be added in the future.
 
 | Algorithm | Gyroscope | Accelerometer | Magnetometer | Status  |
 |-----------|:---------:|:-------------:|:------------:|:-------:|

ahrs/common/quaternion.py

@@ -387,8 +387,6 @@ def slerp(q0: np.ndarray, q1: np.ndarray, t_array: np.ndarray, threshold: float
 
 class Quaternion(np.ndarray):
     """
-    Quaternion.
-
     Representation of a quaternion. It can be built with 3- or 4-dimensional
     vectors. The quaternion object is always normalized to represent rotations
     in 3D space, also known as a **versor**.

ahrs/filters/madgwick.py

@@ -28,7 +28,7 @@
 Orientation from angular rate
 -----------------------------
 
-The orientation of the Earth frame realtive to the sensor frame
+The orientation of the Earth frame relative to the sensor frame
 :math:`\,\\mathbf{q}_{\\omega, t}=\\begin{bmatrix}q_w & q_x & q_y & q_z\\end{bmatrix}`
 at time :math:`t` can be computed by numerically integrating the quaternion
 derivative :math:`\\dot{\\mathbf{q}}_t=\\frac{1}{2}\\mathbf{q}_{t-1}\\mathbf{\\omega}_t` as:
@@ -54,8 +54,8 @@
 is calculated from angular rates.
 
 A more detailed explanation of the orientation estimation solely based on
-angular rate can found in the documentation of the `AngularRate <./angular.html>`_
-estimator.
+angular rate can be found in the documentation of the `AngularRate
+<./angular.html>`_ estimator.
 
 Orientation as solution of Gradient Descent
 -------------------------------------------
@@ -68,7 +68,7 @@
 its corresponding *measured* direction in the sensor frame,
 :math:`^S\\mathbf{s}=\\begin{bmatrix}0 & s_x & s_y & s_z\\end{bmatrix}`.
 
-Thus, the **objective function** is:
+Thus, the `objective function <https://en.wikipedia.org/wiki/Loss_function>` is:
 
 .. math::
     \\begin{array}{rcl}
@@ -363,7 +363,7 @@
 
 References
 ----------
-.. [Madgwick] Sebastian Madgwick. An efficient orientation filter for inertial 
+.. [Madgwick] Sebastian Madgwick. An efficient orientation filter for inertial
     and inertial/magnetic sensor arrays. April 30, 2010.
     http://www.x-io.co.uk/open-source-imu-and-ahrs-algorithms/
 
@@ -373,7 +373,8 @@
 from ..common.orientation import q_prod, q_conj, acc2q, am2q
 
 class Madgwick:
-    """Madgwick's Gradient Descent Orientation Filter
+    """
+    Madgwick's Gradient Descent Orientation Filter
 
     If ``acc`` and ``gyr`` are given as parameters, the orientations will be
     immediately computed with method ``updateIMU``.
@@ -426,8 +427,12 @@ class Madgwick:
     Examples
     --------
     Assuming we have 3-axis sensor data in N-by-3 arrays, we can simply give
-    these samples to their corresponding type. The Madgwick algorithm can work
-    solely with gyroscope and accelerometer samples.
+    these samples to their corresponding type.
+
+    **IMU Array**
+
+    The Madgwick algorithm can work solely with gyroscope and accelerometer
+    samples.
 
     The easiest way is to directly give the full array of samples to their
     matching parameters.
@@ -448,6 +453,8 @@ class Madgwick:
     >>> for t in range(1, num_samples):
     ...     Q[t] = madgwick.updateIMU(Q[t-1], gyr=gyro_data[t], acc=acc_data[t])
 
+    **MARG Array**
+
     Further on, we can also use magnetometer data.
 
     >>> madgwick = Madgwick(gyr=gyro_data, acc=acc_data, mag=mag_data)   # Using MARG
@@ -506,7 +513,8 @@ def __init__(self, gyr: np.ndarray = None, acc: np.ndarray = None, mag: np.ndarr
             self.Q = self._compute_all()
 
     def _compute_all(self) -> np.ndarray:
-        """Estimate the quaternions given all data.
+        """
+        Estimate the quaternions given all data.
 
         Attributes ``gyr`` and ``acc`` must contain data. If ``mag`` contains
         data, the updateMARG() method is used.