forked from aFuerst/PyCUDA-Raster
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathSetup-Instructions.txt
129 lines (98 loc) · 5.59 KB
/
Setup-Instructions.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
Windows (No QGIS):
1) Install Python 2.7.12 from here:
https://www.python.org/downloads/
Add "C:\Python27;" and "C:\Python27\Scripts;" to your system PATH variable
If you have multiple installations of Python on your machine you must ensure
to place these two at the front / top of your PATH.
You MUST install the 64 bit version of Python or the program will NOT work.
2) Install Visual Studio 2013 Community Edition
https://www.visualstudio.com/en-us/news/vs2013-community-vs.aspx
Add these 3 lines to system PATH
"C:\Program Files (x86)\Microsoft Visual Studio 12.0\VC\bin;"
"C:\Program Files (x86)\Microsoft Visual Studio 12.0\VC\bin\amd64;"
"C:\Program Files (x86)\Microsoft Visual Studio 12.0\Common7\IDE;"
You will have to close and re-open any command prompt windows you have open to be able to
do later Python library installations.
3) Install CUDA
https://developer.nvidia.com/cuda-downloads/
4) Install Numpy
http://www.numpy.org/
or run "pip install numpy" in command prompt
5) Install Boost
Install intsructions are located here:
http://www.boost.org/doc/libs/1_61_0/more/getting_started/windows.html
6) Install Boost.Python
http://www.lfd.uci.edu/~gohlke/pythonlibs/#boost.python
Select the one titled "boost_python-1.59-cp27-none-win_amd64.whl"
Install it using "pip install boost_python-1.59-cp27-none-win_amd64.whl"
7) Install Python GDAl
http://www.lfd.uci.edu/~gohlke/pythonlibs/#gdal
Select the one titled "GDAL-2.0.3-cp27-cp27m-win_amd64.whl"
Install using "pip install GDAL-2.0.3-cp27-cp27m-win_amd64.whl"
8) Install PyCUDA
This is a list of PyCUDA binaries:
http://www.lfd.uci.edu/~gohlke/pythonlibs/#pycuda
Select the one titled "pycuda-2016.1.2+cuda7518-cp27-cp27m-win_amd64.whl"
Navigate to your downloads folder in command prompt and run "pip install pycuda-2016.1.2+cuda7518-cp27-cp27m-win_amd64.whl"
9) (OPTIONAL) Install PyQT4
In order to use the standalone GUI, you need the PyQT4 libraries.
http://www.lfd.uci.edu/~gohlke/pythonlibs/#pyqt4
Select the package titles "PyQt4-4.11.4-cp27-none-win_amd64.whl"
Navigate to your downloads folder in command prompt and run "pip install PyQt4-4.11.4-cp27-none-win_amd64.whl"
**THIS INSTALLATION WILL NOT WORK WITH QGIS. OUR PLUGIN DOESN'T CURRENTLY SUPPORT WINDOWS**
##################################################################################
Debian (Ubunutu, Mint, ...) based systems:
1. Make sure Python 2.7 is installed, it probably is by default. QGIS wouldn't
build without a Python installation, so you likely have it.
------------------------------------------------------------------------------
2. Make sure Numpy is installed. This is a QGIS dependency, so it should
already be installed. To make sure Numpy is installed, open the Python
interpreter or write a Python script and type "import numpy". If this fails,
you will need to install Numpy. On Ubuntu you can do this on Ubuntu by using
the following command on the command line.
sudo apt-get install python-numpy
Or, you can use pip which is included with Python.
pip install numpy
------------------------------------------------------------------------------
3. Make sure GDAL for Python is installed. This is another QGIS dependency, so
it should already be installed if you have QGIS. To check if GDAL is
installed, open the Python interpreter or write a Python script and type
"import osgeo". If this fails, you will need to install GDAL. You can do this
on Ubuntu through
sudo apt-get install python-gdal
or, you can download and install GDAL directly. You will need to build GDAL,
and then build the python library on top of it.
http://www.gdal.org/
https://pypi.python.org/pypi/GDAL/
------------------------------------------------------------------------------
4. You will need a compatible NVIDIA driver for your version of Linux, as
well as CUDA 6.0 or greater. Detailed instructions on downloading and
installing CUDA can be found here
http://docs.nvidia.com/cuda/cuda-getting-started-guide-for-linux/index.html
------------------------------------------------------------------------------
5. For PyCUDA to be able to interact with CUDA properly, Boost.Python, part of
Boost C++ libraries must be installed. If Boost isn't already installed, on
Ubuntu you can use
sudo apt-get install libboost-all-dev
or download and install directly from the boost website.
http://www.boost.org/
------------------------------------------------------------------------------
6. Finally, install PyCUDA. The most reliable way to install PyCUDA is to
download it from the Python Package Index and install it by following the
instructions in README_SETUP.txt inside unpackaged folder.
https://pypi.python.org/pypi/pycuda
Or you can try to install it using pip
pip install pycuda
------------------------------------------------------------------------------
7. In order for PyCUDA to recognize the CUDA installation, you many need to
change some path variables. Edit your .bashrc file (usually located in
/home/username/) and add the following lines
export PATH=$PATH:/path/to/cuda/bin
export LD_LIBRARY_PATH=/path/to/cuda/lib64
"/path/to/cuda" should be wherever you have your CUDA installation from step 4.
For me that is "/usr/local/cuda-7.0". Once you're done editing the .bashrc file,
run the command "source .bashrc" to update the changes.
------------------------------------------------------------------------------
8. You should now be able to run the plugin from the command line. In order for
QGIS to properly recognize all of the changes, you may need to restart your
computer.