This is a library to calculate a grid (or multiple grids) of MESA models in an easy way using a Python implementation. It is easy to use as standalone routine using a FORTRAN namelist input file as well as part of an extended code. Currently MESAPC does not work with latest MESA release (15140) but should work with recent ones before that.
MESA is a one dimensional stellar evolution code widely used in stellar astrophysics (For more information please take a look at the intrument papers: MESA I, MESA II, MESA III, MESA IV and MESA V). To run MESAPC you will need a working version of MESA. The installation of MESA and the corresponding MESA SDK is straightforward. A guide can be be found here.
MESAPC makes use of of mulitple Python Libraries. To have this working properly, we strongly recommend working with a virtual environment. MESAPC has been developed in a Python 3.7 environment using the following packages:
- f90nml 1.1.2
- numpy 1.18.2
- tqdm 4.45.0
Older versions of these modules have not been tested. We don' know if they will work.
If you have all the prerequisites installed, just install MESAPC with
pip install MESAPCand set the environmet variable MESAPC_DIR to the directory you are ready to go. If you cloned the repository in i.e. /home/user/Projects then set the environment variable using
export MESAPC_DIR=/home/user/Projects/MESAPCTo run MESAPC activate your environment. Then just use
python -m MESAPCSimilar to MESA, the input needed is a FORTRAN namelist file. It looks for a file called inlist_MESAPC. The basic structure is the following:
&basics
output_dir = "wherever you want the output to be saved"
/
&grid
initial_mass = "linspace,1,10,100"
/
The namelist <basics> defines the global parameters for the grid. At least you will have to specify the output directory. For a list of all options see section "basics defaults".
The namelist grid describes the paramaters on which to calculate the grid. The grid will be sampled over all inputs in the grid namelist. The given example is to calculate a grid of 100 stellar models with linearly spaced masses between 1 and 10 solar masses. Any <star_job> or <controls> input that is allowed for MESA calculations can be used here. MESAPC will check internally if the inputs entered are valid for the MESA version stored ad $MESA_DIR. You have different options to enter values. The input has to be a string in the following form (all values are separated by a comma ,):
- type declaration
- for a list of strings write: string
- for a list of int write: int
- for a list of float write: float
- for a linearly spaced list of float values use: linspace (numpy linspace is used to create the list)
- COMING SOON: for a logarithmic spaced list of float values use: logspace
- input values
- for a list just type the values separated with a comma
- for a linspace list type the elements of np.linspace(a, b, c) in the order a,b,c (again separated by a comma)
- for a logspace list type the elements of 10**np.linspace(log10(a), log10(b), c) in the order a,b,c
Examples can be found in filebase/example_inputs.
You can add any number of additional grid runs. These may be seen in the following way. The grid is equal to some standard model. If you then want to look at changes to the standard model you can add additional runs with i.e.
&run1
initial_zfracs = "int,4,5,6,7,8"
/
&run2
mixing_length_alpha = "f,1,1.5,2"
/
The additional runs will be sampled with the normal grid and the run inputs are added to the sampling. run1 will therefore result in the calculation of 100 models with linearly spaced masses in the range of 1 to 10 solar masses, each calculated for all initial_zfracs values 4 5 6 7 8. Consequently run2 will calculate all masses for the different values of mixing length. To achieve a combination of this (all masses with all initial_zfracs and all mixing_length alpha), the input from run2 would have to be added to the input of run1.
MESAPC caluclates the grids as defined in inlist_MESAPC. During the evolution you will see a tqdm progress bar for each run. After the evolutionas have finished the output_dir will hold the following:
- inlist_project: The initial inlist project input file
- LOGS: a folder full of more folders. Each of the folders (called grid_1, grid_2, ...grid_N, run1_1 run1_2, ..., run1_M, run2_1, etc.) will have the elements that would typically be in by LOGS directory in a MESA folder. This means the history.data and profile.data files as specified in the inlist_project (Output might be changed for different runs by including this in the
<run>namelists as lists with one entry). In addition, the folder will hold two textfiles showing what is different in this calculation in respect to the inlist_project file. One for the star_job namelist and one for the controls inlist. - Running
python -m MESAPC trackershows a HR diagramm of the grid (and runs). This can be used to follow the progress of the calculations.
The <basic> namelist has the follwing options:
- NUM_THREADS: Defines the number of threads used for a single MESA calculation (default 2)
- TOTAL_NUM_THREADS: Defines the total number of threads MESAPC can use. Together with NUM_THREADS, this defines how much calculation can be run in parallel. (default 10)
- calculate_grid: If True then calculate the grid. If false, only the runs will be calculated (default True)
- inlist: The default inlist is given in filebase/inlist_project. It creates a pre-main-sequence model and calculates the evolution until the ZAMS (central hydrogen below 0.68). To change the inlist, just give the full path to your inlist.
Using controls that work with vectors in grids introduces problems. Currently, only grids work properly for vectors. We will show how to work with them with an example: When using extra controls, e.g. x_ctrl(:), make sure to use all of them in the namelist, even if they have only one value. I you want to calculate the grid for x_ctrl(3) = 1.0, 2.0, 3.0, 4.4, your inlist_MESAPC should look something like:
&basics
output_dir = "change_x_ctrl(3)"
/
&grid
x_ctrl(1) = "float,3.14"
x_ctrl(2) = "float,0.01"
x_ctrl(3) = "float,1.0,2.0,3.0,4.0
/