QuickStart¶
Note
On Windows, 32 bit python 3.8 and above are not supported.
For a Windows install, jump straight to Windows Batch File
Install Numpy & Matplotlib¶
Because RocketCEA depends on f2py to compile the FORTRAN NASA CEA code and f2py is part of numpy , numpy will need to be installed.
RocketCEA also makes use of the matplotlib package for creating plots.
To install numpy and matplotlib, give the commands:
pip install numpy
pip install matplotlib
OR perhaps...
pip install --upgrade numpy
pip install --upgrade matplotlib
Some Linux systems may require:
sudo pip install numpy
sudo pip install matplotlib
Install Compiler¶
Using f2py to compile FORTRAN requires a FORTRAN compiler. I recommend using gfortran on all platforms so that there are no FORTRAN incompatibilities between platforms.
Each operating system has its own approach to install gfortran
Click: Install gfortran to see install instructions for a few platforms that I have tested.
Important
Windows users MUST put MinGW into environment PATH variable. (see: Windows PATH)
for 32 bit: C:\MinGW\mingw32\bin AND C:\MinGW\mingw32\bin
For 64 bit: C:\MinGW\mingw64\lib AND C:\MinGW\mingw64\lib
Install RocketCEA¶
After the above installs have been accomplished, the easiest way to install RocketCEA is:
pip install rocketcea
OR on Linux
sudo pip install rocketcea
OR perhaps
pip install --user rocketcea
Try a quick test of the install by pasting the following into a command terminal:
python -c "from rocketcea.cea_obj import CEA_Obj; C=CEA_Obj(oxName='LOX', fuelName='LH2'); print(C.get_Isp())"
should result in:
374.30361765576265
Windows Batch File¶
Note
for Anaconda on Windows, see “Anaconda Windows Batch File” section below.
RocketCEA on Windows can be problematic. Often the problem is in the PATH environment variable where the wrong files are found for the intended install, and even more often it is because the proper compiler build tools are not available.
The Windows batch file below addresses both of those problems.
Remember that the FORTRAN compiler must also be installed (see: Install MinGW on Windows 10)
Notice that the batch file makes the PATH as simple as possible so that only the 64 bit MinGW files and desired python files are found.
Note that the batch file assumes that python 3.9 64 bit is the python version installed at C:\Python39_64 and that MinGW 64 bit is installed at C:\MinGW\mingw64\bin and C:\MinGW\mingw64\lib. Edit those path names for your situation.
Note that it starts by uninstalling rocketcea in case bad files are left from previous attempts. (Plan for the worst, hope for the best.)
You may need to edit the hard-coded paths to your own location of python and MinGW.
- Copy and paste the batch file code below into an editor
- Edit the hard-coded paths to your own location of python and MinGW (i.e. perhaps change C:\Python39_64 and C:\MinGW\mingw64)
- Save the edited BAT file (e.g. as RUN_SETUP_BUILD_WIN64.BAT)
- Open a command prompt terminal and navigate to the BAT file directory.
- Give the command RUN_SETUP_BUILD_WIN64.BAT
Note
9/16/2021: BAT File successful on Windows 10 with 64 bit python 3.8.10 and 3.9.7
rem =============== RUN_SETUP_BUILD_WIN64.BAT ================
SET "MYPYTHONPATH=C:\Python39_64"
rem Make sure that PATH is as simple as possible
set PATH=C:\MinGW\mingw64\bin;C:\MinGW\mingw64\lib;%MYPYTHONPATH%;%MYPYTHONPATH%\Scripts
pip uninstall -y rocketcea
pip install future
pip install numpy
pip install scipy
pip install pillow
pip install matplotlib
pip install --global-option build_ext --global-option --compiler=mingw32 rocketcea
rem Test the compiled module
python -c "from rocketcea.cea_obj import CEA_Obj; C=CEA_Obj(oxName='LOX', fuelName='LH2'); print(C.get_Isp())"
Note
The “trick” for installing RocketCEA on Windows seems to be pre-installing the dependency packages (e.g. numpy, scipy and matplotlib) so that the pip options for rocketcea only apply to rocketcea.
Anaconda Windows Batch File¶
As stated above, the “trick” for installing RocketCEA on Windows seems to be pre-installing the dependency packages (e.g. numpy, scipy and matplotlib) so that the pip options for rocketcea only apply to rocketcea.
On Anaconda, many of the required dependencies are preinstalled.
To test the required installs, open an Anaconda prompt, launch python and attempt to import each package. (An example is shown below for Anaconda with python 3.9)
Notice that a fresh install of Anaconda does not include the pillow package.
use the command:
conda install -c anaconda pillow
to install pillow . The results will look something like the image below.
With all the dependencies handled, the batch file can be much more simple than the one shown above. Before you can run it, you need to verify the location of the python folders under Anaconda.
To find your python folder, open an Anaconda prompt, launch python, import sys and give the command “sys.executable” as in the image below.
In my case, the path was:
C:\Users\Charlie\anaconda3
Yours will likely be similar.
For the final step, edit the batch file below to replace <YourUserName> with the user name shown in the python path above.
rem NOTICE: MUST Verify dependency installs prior to running this batch file.
SET "MYPYTHONPATH=C:\Users\<YourUserName>\anaconda3"
rem Make sure that PATH is as simple as possible... For Anaconda notice "\Libary\bin"
set PATH=C:\MinGW\mingw64\bin;C:\MinGW\mingw64\lib;%MYPYTHONPATH%;%MYPYTHONPATH%\Scripts;%MYPYTHONPATH%\Library\bin
pip install --global-option build_ext --global-option --compiler=mingw32 rocketcea
rem Test the compiled module
python -c "from rocketcea.cea_obj import CEA_Obj; C=CEA_Obj(oxName='LOX', fuelName='LH2'); print(C.get_Isp())"
Run the batch file and the results will hopefully be similar to the results shown below.
Notice that the file ran the “quick test” at the bottom of the batch file and properly output the Isp of 374.30…
Getting Help¶
After installing with pip, there should be a launch command line program called rocketcea or, on Windows, rocketcea.exe.
From a terminal or command prompt window simply type:
rocketcea
Your browser will launch with these RocketCEA help pages.
Running RocketCEA¶
Any use of RocketCEA begins with an import statement and an instance of CEA_obj:
from rocketcea.cea_obj import CEA_Obj
C = CEA_Obj( oxName='LOX', fuelName='LH2')
In the above example, LOX and LH2 are called out, but any propellants on the Propellants page can be used.
There are a large number of examples included in this document
For instance, look at LOX/LH2 Performance on the Standard Examples page.
To run an example, highlight the source code with your mouse, right click the highlighted code and select Copy. Paste that code into your text editor and save it to a python file.(for example D:\rocketcea\example_1.py).
Example files can be run with the command:
python example1.py
Or, in many text editors hitting the F5 key will execute the code.
Note
RocketCEA is compiled with the mingw and mingw-w64 gfortran compilers using default f2py options giving a “shared” *.pyd file that requires mingw libraries at run time.
If you see the error: Import Error: DLL load failed: The specified module could not be found
You may need to install the MinGW Compiler Suite and perhaps even recompile RocketCEA in order
for RocketCEA to work (see below)
On Windows, make sure the environment PATH variable is set properly (see: Windows PATH)
Test The Install¶
Paste the following code into your text editor and save it to your test folder as basic_cea.py (for example, D:\rocketcea\basic_cea.py):
from rocketcea.cea_obj import CEA_Obj
C = CEA_Obj( oxName='LOX', fuelName='LH2')
for mr in range(2,9):
print(mr, C.get_Isp(Pc=100.0, MR=mr, eps=40.0) )
At the command prompt, give the command:
python basic_cea.py
If you see the following output:
(2, 424.3597085736007)
(3, 445.44434236555196)
(4, 453.13271951921837)
(5, 453.240429182719)
(6, 448.190232998362)
(7, 438.74340042907266)
(8, 424.6998266323161)
Great… you are good to go.
If not, see the information below.
Google Colaboratory¶
If you are having trouble installing RocketCEA on your system, RocketCEA can be run on Google Colaboratory (either python3 or python2).
Colaboratory is a free Jupyter notebook environment that requires no setup and runs entirely in the cloud.
After creating a Colaboratory notebook, install RocketCEA.:
!pip install RocketCEA
If Needed, install libgfortran3:
!apt-get install libgfortran3
Create a python script to run RocketCEA:
%%file chk_cea.py
from rocketcea.cea_obj import CEA_Obj
C = CEA_Obj( oxName='LOX', fuelName='LH2')
for mr in range(2,9):
print(mr, C.get_Isp(Pc=100.0, MR=mr, eps=40.0) )
And then run the file:
!python chk_cea.py
Colab plots work with RocketCEA as well.
Windows 10 with WSL¶
RocketCEA can also be installed on Windows Subsystem for Linux (WSL)
After setting up your Linux distribution on WSL, installing RocketCEA is quick and easy.
For example on a Ubuntu distribution.
- Update Ubuntu
- sudo apt-get update
- configure Ubuntu for python3 and FORTRAN
- sudo apt install python3-pip
- sudo apt-get install gfortran
- Install RocketCEA required libraries
- pip install numpy
- pip install matplotlib
- pip install rocketcea
- Test installation with quick example
- python3 -c “from rocketcea.cea_obj import CEA_Obj; C=CEA_Obj(oxName=’LOX’, fuelName=’LH2’); print(C.get_Isp())”
- SHOULD RETURN: 374.30361765576896 (or perhaps just very close to it.)