updated docs
This commit is contained in:
Zoltán Vörös
parent
8e16ab25e5
commit
208fa196ab
17 changed files with 109 additions and 110 deletions
|
|
@ -27,7 +27,7 @@ copyright = '2019-2021, Zoltán Vörös and contributors'
|
|||
author = 'Zoltán Vörös'
|
||||
|
||||
# The full version, including alpha/beta/rc tags
|
||||
release = '2.1.1'
|
||||
release = '2.1.2'
|
||||
|
||||
|
||||
# -- General configuration ---------------------------------------------------
|
||||
|
|
|
|||
|
|
@ -14,7 +14,7 @@ Welcome to the ulab book!
|
|||
ulab-intro
|
||||
|
||||
.. toctree::
|
||||
:maxdepth: 3
|
||||
:maxdepth: 2
|
||||
:caption: User's guide:
|
||||
|
||||
ulab-ndarray
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
None
|
||||
|
||||
Fourier transforms
|
||||
==================
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
None
|
||||
|
||||
Numpy functions
|
||||
===============
|
||||
|
||||
|
|
@ -20,7 +20,7 @@ from ``numpy``.
|
|||
13. `numpy.median <#median>`__
|
||||
14. `numpy.min <#min>`__
|
||||
15. `numpy.minimum <#minimum>`__
|
||||
16. `numpy.not_equal <#not_equal>`__
|
||||
16. `numpy.not_equal <#equal>`__
|
||||
17. `numpy.polyfit <#polyfit>`__
|
||||
18. `numpy.polyval <#polyval>`__
|
||||
19. `numpy.roll <#roll>`__
|
||||
|
|
@ -570,22 +570,13 @@ https://docs.scipy.org/doc/numpy/reference/generated/numpy.min.html
|
|||
|
||||
See `numpy.max <#max>`__.
|
||||
|
||||
.. code::
|
||||
|
||||
# code to be run in CPython
|
||||
|
||||
|
||||
minimum
|
||||
-------
|
||||
|
||||
``numpy``:
|
||||
https://docs.scipy.org/doc/numpy/reference/generated/numpy.minimum.html
|
||||
|
||||
Returns the minimum of two arrays, or two scalars, or an array, and a
|
||||
scalar. If the arrays are of different ``dtype``, the output is upcast
|
||||
as in `Binary operators <#Binary-operators>`__. If both inputs are
|
||||
scalars, a scalar is returned. Only positional arguments are
|
||||
implemented.
|
||||
See `numpy.maximum <#maximum>`__
|
||||
|
||||
maximum
|
||||
-------
|
||||
|
|
@ -1045,8 +1036,3 @@ points corresponding to ``y``.
|
|||
|
||||
|
||||
|
||||
|
||||
.. code::
|
||||
|
||||
# code to be run in CPython
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
None
|
||||
|
||||
Linalg
|
||||
======
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
None
|
||||
|
||||
Universal functions
|
||||
===================
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
None
|
||||
|
||||
Optimize
|
||||
========
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
None
|
||||
|
||||
Signal
|
||||
======
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
None
|
||||
|
||||
Special functions
|
||||
=================
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
None
|
||||
|
||||
Introduction
|
||||
============
|
||||
|
||||
|
|
@ -101,7 +101,7 @@ approximately 120 kB of extra compiled code to the ``micropython``
|
|||
can easily shave tens of kB off the firmware. In fact, if only a small
|
||||
sub-set of functions are needed, you can get away with less than 10 kB
|
||||
of flash space. See the section on `customising
|
||||
ulab <#Custom_builds>`__.
|
||||
ulab <#Customising-the-firmware>`__.
|
||||
|
||||
Resources and legal matters
|
||||
---------------------------
|
||||
|
|
@ -145,13 +145,13 @@ implementation details only, which all have been sorted out with the
|
|||
generous and enthusiastic support of Jeff Epler from `Adafruit
|
||||
Industries <http://www.adafruit.com>`__.
|
||||
|
||||
There are, however, a couple of instances, where the usage in the two
|
||||
environments is different at the python level. These are how the class
|
||||
properties can be accessed. We will point out the differences and
|
||||
possible workarounds at the relevant places in this document.
|
||||
There are, however, a couple of instances, where the two environments
|
||||
differ at the python level in how the class properties can be accessed.
|
||||
We will point out the differences and possible workarounds at the
|
||||
relevant places in this document.
|
||||
|
||||
Customising ``ulab``
|
||||
====================
|
||||
Customising the firmware
|
||||
========================
|
||||
|
||||
As mentioned above, ``ulab`` has considerably grown since its
|
||||
conception, which also means that it might no longer fit on the
|
||||
|
|
@ -587,8 +587,3 @@ the
|
|||
#define NDARRAY_HAS_BINARY_OP_POWER (0)
|
||||
|
||||
definition.
|
||||
|
||||
.. code::
|
||||
|
||||
# code to be run in CPython
|
||||
|
||||
|
|
|
|||
|
|
@ -1,6 +1,6 @@
|
|||
None
|
||||
ndarray, the basic container
|
||||
============================
|
||||
|
||||
ndarray, the base class
|
||||
=======================
|
||||
|
||||
The ``ndarray`` is the underlying container of numerical data. It can be
|
||||
thought of as micropython’s own ``array`` object, but has a great number
|
||||
|
|
@ -250,7 +250,17 @@ type, or performance is important.
|
|||
Array initialisation functions
|
||||
------------------------------
|
||||
|
||||
There are seven functions that can be used for initialising an array.
|
||||
There are nine functions that can be used for initialising an array.
|
||||
|
||||
1. `numpy.arange <#arange>`__
|
||||
2. `numpy.concatenate <#concatenate>`__
|
||||
3. `numpy.eye <#eye>`__
|
||||
4. `numpy.frombuffer <#frombuffer>`__
|
||||
5. `numpy.full <#full>`__
|
||||
6. `numpy.linspace <#linspace>`__
|
||||
7. `numpy.logspace <#logspace>`__
|
||||
8. `numpy.ones <#ones>`__
|
||||
9. `numpy.zeros <#zeros>`__
|
||||
|
||||
arange
|
||||
~~~~~~
|
||||
|
|
@ -442,6 +452,52 @@ Specifying the dimensions of the matrix
|
|||
|
||||
|
||||
|
||||
frombuffer
|
||||
~~~~~~~~~~
|
||||
|
||||
``numpy``:
|
||||
https://numpy.org/doc/stable/reference/generated/numpy.frombuffer.html
|
||||
|
||||
The function interprets a contiguous buffer as a one-dimensional array,
|
||||
and thus can be used for piping buffered data directly into an array.
|
||||
This method of analysing, e.g., ADC data is much more efficient than
|
||||
passing the ADC buffer into the ``array`` constructor, because
|
||||
``frombuffer`` simply creates the ``ndarray`` header and blindly copies
|
||||
the memory segment, without inspecting the underlying data.
|
||||
|
||||
The function takes a single positional argument, the buffer, and three
|
||||
keyword arguments. These are the ``dtype`` with a default value of
|
||||
``float``, the ``offset``, with a default of 0, and the ``count``, with
|
||||
a default of -1, meaning that all data are taken in.
|
||||
|
||||
.. code::
|
||||
|
||||
# code to be run in micropython
|
||||
|
||||
from ulab import numpy as np
|
||||
|
||||
buffer = b'\x01\x02\x03\x04\x05\x06\x07\x08'
|
||||
print('buffer: ', buffer)
|
||||
|
||||
a = np.frombuffer(buffer, dtype=np.uint8)
|
||||
print('a, all data read: ', a)
|
||||
|
||||
b = np.frombuffer(buffer, dtype=np.uint8, offset=2)
|
||||
print('b, all data with an offset: ', b)
|
||||
|
||||
c = np.frombuffer(buffer, dtype=np.uint8, offset=2, count=3)
|
||||
print('c, only 3 items with an offset: ', c)
|
||||
|
||||
.. parsed-literal::
|
||||
|
||||
buffer: b'\x01\x02\x03\x04\x05\x06\x07\x08'
|
||||
a, all data read: array([1, 2, 3, 4, 5, 6, 7, 8], dtype=uint8)
|
||||
b, all data with an offset: array([3, 4, 5, 6, 7, 8], dtype=uint8)
|
||||
c, only 3 items with an offset: array([3, 4, 5], dtype=uint8)
|
||||
|
||||
|
||||
|
||||
|
||||
full
|
||||
~~~~
|
||||
|
||||
|
|
@ -2285,8 +2341,3 @@ view is really nothing but a short header with a data array that already
|
|||
exists, and is filled up. Hence, creating the view requires only the
|
||||
creation of its header. This operation is fast, and uses virtually no
|
||||
RAM.
|
||||
|
||||
.. code::
|
||||
|
||||
# code to be run in CPython
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
None
|
||||
|
||||
Programming ulab
|
||||
================
|
||||
|
||||
|
|
@ -909,8 +909,3 @@ function are
|
|||
MP_DEFINE_CONST_FUN_OBJ_N()
|
||||
3. Binding this function object to the namespace in the
|
||||
``ulab_user_globals_table[]``
|
||||
|
||||
.. code::
|
||||
|
||||
# code to be run in CPython
|
||||
|
||||
|
|
|
|||
|
|
@ -247,7 +247,7 @@
|
|||
"1. [numpy.median](#median)\n",
|
||||
"1. [numpy.min](#min)\n",
|
||||
"1. [numpy.minimum](#minimum)\n",
|
||||
"1. [numpy.not_equal](#not_equal)\n",
|
||||
"1. [numpy.not_equal](#equal)\n",
|
||||
"1. [numpy.polyfit](#polyfit)\n",
|
||||
"1. [numpy.polyval](#polyval)\n",
|
||||
"1. [numpy.roll](#roll)\n",
|
||||
|
|
@ -916,13 +916,6 @@
|
|||
"See [numpy.max](#max)."
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": []
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
|
|
@ -931,8 +924,13 @@
|
|||
"\n",
|
||||
"`numpy`: https://docs.scipy.org/doc/numpy/reference/generated/numpy.minimum.html\n",
|
||||
"\n",
|
||||
"Returns the minimum of two arrays, or two scalars, or an array, and a scalar. If the arrays are of different `dtype`, the output is upcast as in [Binary operators](#Binary-operators). If both inputs are scalars, a scalar is returned. Only positional arguments are implemented.\n",
|
||||
"\n",
|
||||
"See [numpy.maximum](#maximum)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"## maximum\n",
|
||||
"\n",
|
||||
"`numpy`: https://docs.scipy.org/doc/numpy/reference/generated/numpy.maximum.html\n",
|
||||
|
|
@ -1520,13 +1518,6 @@
|
|||
"print('integral of y: ', np.trapz(y))\n",
|
||||
"print('integral of y at x: ', np.trapz(y, x=x))"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": []
|
||||
}
|
||||
],
|
||||
"metadata": {
|
||||
|
|
|
|||
|
|
@ -14,11 +14,11 @@
|
|||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 7,
|
||||
"execution_count": 1,
|
||||
"metadata": {
|
||||
"ExecuteTime": {
|
||||
"end_time": "2021-01-14T17:57:42.304478Z",
|
||||
"start_time": "2021-01-14T17:57:42.205232Z"
|
||||
"end_time": "2021-01-15T13:53:42.464150Z",
|
||||
"start_time": "2021-01-15T13:53:42.449894Z"
|
||||
}
|
||||
},
|
||||
"outputs": [
|
||||
|
|
@ -61,7 +61,7 @@
|
|||
"author = 'Zoltán Vörös'\n",
|
||||
"\n",
|
||||
"# The full version, including alpha/beta/rc tags\n",
|
||||
"release = '2.1.1'\n",
|
||||
"release = '2.1.2'\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"# -- General configuration ---------------------------------------------------\n",
|
||||
|
|
@ -148,11 +148,11 @@
|
|||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 2,
|
||||
"execution_count": 35,
|
||||
"metadata": {
|
||||
"ExecuteTime": {
|
||||
"end_time": "2021-01-14T16:17:06.823520Z",
|
||||
"start_time": "2021-01-14T16:17:06.759683Z"
|
||||
"end_time": "2021-01-15T14:18:49.025168Z",
|
||||
"start_time": "2021-01-15T14:18:49.015858Z"
|
||||
}
|
||||
},
|
||||
"outputs": [
|
||||
|
|
@ -182,7 +182,7 @@
|
|||
" ulab-intro\n",
|
||||
"\n",
|
||||
".. toctree::\n",
|
||||
" :maxdepth: 3\n",
|
||||
" :maxdepth: 2\n",
|
||||
" :caption: User's guide:\n",
|
||||
"\n",
|
||||
" ulab-ndarray\n",
|
||||
|
|
@ -212,11 +212,11 @@
|
|||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 4,
|
||||
"execution_count": 31,
|
||||
"metadata": {
|
||||
"ExecuteTime": {
|
||||
"end_time": "2021-01-14T16:19:01.938959Z",
|
||||
"start_time": "2021-01-14T16:18:59.040730Z"
|
||||
"end_time": "2021-01-15T14:09:47.022621Z",
|
||||
"start_time": "2021-01-15T14:09:46.985214Z"
|
||||
}
|
||||
},
|
||||
"outputs": [],
|
||||
|
|
@ -234,6 +234,7 @@
|
|||
"def convert_notebook(fn):\n",
|
||||
" source = nb.read(fn+'.ipynb', nb.NO_CONVERT)\n",
|
||||
" notebook = nb4.new_notebook()\n",
|
||||
" notebook.cells = []\n",
|
||||
" append_cell = False\n",
|
||||
" for cell in source['cells']:\n",
|
||||
" if append_cell:\n",
|
||||
|
|
@ -245,16 +246,17 @@
|
|||
" \n",
|
||||
" (rst, resources) = rstexporter.from_notebook_node(notebook)\n",
|
||||
" with open('./manual/source/' + fn + '.rst', 'w') as fout:\n",
|
||||
" # it's a bit odd, but even an emtpy notebook is converted into a \"None\" string\n",
|
||||
" rst = rst.lstrip('None')\n",
|
||||
" fout.write(rst)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 5,
|
||||
"execution_count": null,
|
||||
"metadata": {
|
||||
"ExecuteTime": {
|
||||
"end_time": "2021-01-14T16:19:15.242085Z",
|
||||
"start_time": "2021-01-14T16:19:01.955650Z"
|
||||
"start_time": "2021-01-15T14:38:15.993Z"
|
||||
}
|
||||
},
|
||||
"outputs": [],
|
||||
|
|
|
|||
|
|
@ -189,7 +189,7 @@
|
|||
"%%micropython -pyboard 1\n",
|
||||
"\n",
|
||||
"import utime\n",
|
||||
"import ulab as np\n",
|
||||
"from ulab import numpy as np\n",
|
||||
"\n",
|
||||
"def timeit(n=1000):\n",
|
||||
" def wrapper(f, *args, **kwargs):\n",
|
||||
|
|
@ -277,7 +277,7 @@
|
|||
"- filtering of data (convolution and second-order filters)\n",
|
||||
"- function minimisation, fitting, and numerical approximation routines\n",
|
||||
"\n",
|
||||
"`ulab` implements close to a hundred functions and array methods. At the time of writing this manual (for version 2.1.0), the library adds approximately 120 kB of extra compiled code to the `micropython` (pyboard.v.11) firmware. However, if you are tight with flash space, you can easily shave tens of kB off the firmware. In fact, if only a small sub-set of functions are needed, you can get away with less than 10 kB of flash space. See the section on [customising ulab](#Custom_builds).\n",
|
||||
"`ulab` implements close to a hundred functions and array methods. At the time of writing this manual (for version 2.1.0), the library adds approximately 120 kB of extra compiled code to the `micropython` (pyboard.v.11) firmware. However, if you are tight with flash space, you can easily shave tens of kB off the firmware. In fact, if only a small sub-set of functions are needed, you can get away with less than 10 kB of flash space. See the section on [customising ulab](#Customising-the-firmware).\n",
|
||||
"\n",
|
||||
"## Resources and legal matters\n",
|
||||
"\n",
|
||||
|
|
@ -297,14 +297,14 @@
|
|||
"\n",
|
||||
"`ulab` has originally been developed for `micropython`, but has since been integrated into a number of its flavours. Most of these flavours are simply forks of `micropython` itself, with some additional functionality. One of the notable exceptions is `circuitpython`, which has slightly diverged at the core level, and this has some minor consequences. Some of these concern the C implementation details only, which all have been sorted out with the generous and enthusiastic support of Jeff Epler from [Adafruit Industries](http://www.adafruit.com).\n",
|
||||
"\n",
|
||||
"There are, however, a couple of instances, where the usage in the two environments is different at the python level. These are how the class properties can be accessed. We will point out the differences and possible workarounds at the relevant places in this document."
|
||||
"There are, however, a couple of instances, where the two environments differ at the python level in how the class properties can be accessed. We will point out the differences and possible workarounds at the relevant places in this document."
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"# Customising `ulab`\n",
|
||||
"# Customising the firmware\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"As mentioned above, `ulab` has considerably grown since its conception, which also means that it might no longer fit on the microcontroller of your choice. There are, however, a couple of ways of customising the firmware, and thereby reducing its size. \n",
|
||||
|
|
@ -773,13 +773,6 @@
|
|||
"\n",
|
||||
"definition."
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": []
|
||||
}
|
||||
],
|
||||
"metadata": {
|
||||
|
|
|
|||
|
|
@ -223,7 +223,7 @@
|
|||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"# ndarray, the basic container\n",
|
||||
"# ndarray, the base class\n",
|
||||
"\n",
|
||||
"The `ndarray` is the underlying container of numerical data. It can be thought of as micropython's own `array` object, but has a great number of extra features starting with how it can be initialised, which operations can be done on it, and which functions can accept it as an argument. One important property of an `ndarray` is that it is also a proper `micropython` iterable.\n",
|
||||
"\n",
|
||||
|
|
@ -3396,13 +3396,6 @@
|
|||
"source": [
|
||||
"The fact that the underlying data of a view is the same as that of the original array has another important consequence, namely, that the creation of a view is cheap. Both in terms of RAM, and execution time. A view is really nothing but a short header with a data array that already exists, and is filled up. Hence, creating the view requires only the creation of its header. This operation is fast, and uses virtually no RAM."
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": []
|
||||
}
|
||||
],
|
||||
"metadata": {
|
||||
|
|
|
|||
|
|
@ -730,13 +730,6 @@
|
|||
"2. The definition of a function object by calling MP_DEFINE_CONST_FUN_OBJ_N()\n",
|
||||
"3. Binding this function object to the namespace in the `ulab_user_globals_table[]`"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": []
|
||||
}
|
||||
],
|
||||
"metadata": {
|
||||
|
|
|
|||
Loading…
Reference in a new issue