Update installation for support m2 mac

.github/workflows/ci.yml

@@ -92,9 +92,9 @@ jobs:
       - name: Running pylint
         shell: bash -l {0}
         run: |
+          pylint firecrown
           pylint --rcfile tests/pylintrc  tests
           pylint --rcfile firecrown/models/pylintrc firecrown/models
-          pylint firecrown
       - name: Running pytest
         shell: bash -l {0}
         run: python -m pytest -vv

docs/apple_m1_instructions.rst

@@ -4,92 +4,5 @@ Instructions for installation of Firecrown on Apple M1 and M2 hardware
 .. role:: bash(code)
    :language: bash
 
-The ability to install Firecrown on Apple M1 and M2 hardware (not on Intel hardware) is currently incomplete because of difficulties in obtaining a consistent build of the CosmoSIS Standard Library on M1 and M2 Macs, and the unavailability of an M1 or M2 build of CosmoSIS through conda.
-Until this is fixed, we recommend the following procedure.
-
-.. note::
+Special directions are no longer needed for installation of Firecrown on Apple M1 and M2 hardware.
 
-
-    This procedure is *not* recommended for other platforms.
-    It involves *building* the "head" versions of CosmoSIS and using (parts of) the "head" version of the CosmoSIS Standard Library.
-    The normal installation instructions give you a specified version known to be consistent with the rest of the code.
-    In addition, because this involves *building* CosmoSIS code, if you have something else in your environment that is accidentally found by the build, you may encounter a difficult-to-debug build failure, or worse yet a subtly inconsistent build that will perform incorrectly only some of the time.
-
-Note that this installation procedure will give you an environment in which you have a copy of the Firecrown code and can do development, including producing pull requests to submit code back to Firecrown.
-
-To do development work on Firecrown you need both the Firecrown source code and all the packages upon which Firecrown depends or is used with.
-To get the Firecrown source code you will need to clone the Firecrown repository.
-Most (but not all) of the dependencies of Firecrown are available through conda.
-It is not currently possible to build the CosmoSIS Standard Library on M1 or M2 Macs.
-However, you will be able to use the portions of the CosmoSIS Standard Library that do not require compilation.
-This includes CAMB, but does not include many other modules.
-
-These instructions include details on how to obtain the samplers used with Firecrown.
-This is important because if you are doing development it is important to make sure what you change or add works with both of the supported samplers.
-
-You only need to clone the Firecrown repository once, and to create the conda environment once.
-Every time you want to do development in a new shell session you will need to activate the conda environment and set the environment variables.
-
-Clone the Firecrown repository
-==============================
-
-Choose a directory in which to work.
-In this directory, you will be cloning the Firecrown repository and later building some of the non-Firecrown code that is not installable through conda.
-
-.. code:: bash
-
-    cd /directory/for/firecrown/work
-    git clone https://github.com/LSSTDESC/firecrown.git
-
-
-Installation of dependencies
-============================
-
-These instructions will create a new conda environment containing all the packages used in development.
-This includes testing and code verification tools used during the development process.
-
-It is best to execute these commands while in the same directory as you were in when you cloned the Firecrown repository above.
-Note that we do not build the CosmoSIS Standard Library.
-Only the pure-python parts of the CSL will be available for use.
-
-.. code:: bash
-
-    cd /directory/for/firecrown/work
-    conda create --name firecrown_developer -c conda-forge sacc pyccl fitsio fuzzywuzzy urllib3 PyYAML portalocker idna dill charset-normalizer requests matplotlib flake8 pylint black pytest coverage
-    conda activate firecrown_developer
-    export CC=clang CXX=clang++ FC=gfortran
-    python -m pip install cosmosis cobaya
-    source cosmosis-configure
-    git clone https://github.com/joezuntz/cosmosis-standard-library.git
-
-
-Setting your environment for development
-========================================
-
-Each time you want to do development in a new shell session you need to activate the conda environment and set some environment variables.
-
-Begin by `cd`-ing to the working directory you used during the installation (which should contain both `firecrown` and `cosmosis-standard-library` directories).
-
-.. code:: bash
-
-    cd /directory/for/firecrown/work
-    conda activate firecrown_developer
-    export CSL_DIR=${PWD}/cosmosis-standard-library
-    export FIRECROWN_DIR=${PWD}/firecrown
-    export PYTHONPATH=${FIRECROWN_DIR}/build/lib
-
-To build the Firecrown code you should be in the Firecrown directory:
-
-.. code:: bash
-
-    cd ${FIRECROWN_DIR}
-    python setup.py build
-
-The tests can be run with :bash:`pytest`, after building:
-
-.. code:: bash
-
-    pytest
-
-Examples can be run by `cd`-ing into the specific examples directory and following the instructions in the local README file.
-Note that any example that uses a compiled module from the CosmoSIS Standard Library will fail.

docs/basic_usage.rst

@@ -1,10 +1,8 @@
-
-
 Using Firecrown
 ===============
 
 The documentation for Firecrown is still under development.
 
 There are examples in the `examples folder <https://github.com/LSSTDESC/firecrown/tree/master/examples>`_
-that show some use of Firecrown with both Cobaya and CosmoSIS.
+that show some use of Firecrown with all of the supported samplers.
 

docs/contrib.rst

@@ -1,4 +1,3 @@
-
 Contributing
 ============
 
@@ -7,15 +6,18 @@ Contributing
 
 Contributions to Firecrown are welcome.
 
-For any contribution, please start by `opening an issue <https://github.com/LSSTDESC/firecrown/issues>`_,
-and using the GitHub interface to create a branch for that issue.
+For any contribution, please start by `opening a discussion <https://github.com/LSSTDESC/firecrown/discussions>`_.
+We are intending to use GitHub discussions to come to a consensus on the ideas for new additions.
+Once a consensus is reached, we will convert the discussion into a GitHub issue that can be used to track the progress of the work.
 
+New development for issues should be done on a branch.
 To create a branch you will need write access; if you don't have write access, please send a request to the @LSSTDESC/firecrown-devs team.
 You can also fork the repository and send a pull request from your fork.
 
 When you have completed the task, push your commits to the branch you created for the issue and create a pull request.
 
 We are using several tools to help keep the code tidy and correct; these are described below.
+These tools are all applied by the continuous integration (CI) system that is run on every pull request.
 
 Type checking
 -------------
@@ -33,29 +35,31 @@ and fix any errors reported before pushing commits to the GitHub repository.
 Testing
 -------
 
-
 .. warning::
 
-    We are working on improving the coverage of testing for Firecrown; it is currently very inadequate.
+    We are working on improving the coverage of testing for Firecrown; it is currently inadequate.
     As the coverage improves, we will provide instructions for writing tests for new code.
 
 We are using :bash:`pytest` to run tests on Firecrown.
-Before running tests, one must first build the code.
-In addition, the environment variable :bash:`PYTHONPATH` must be correctly set to run the tests.
-Please see the instructions, above, for this setup.
 
 Use of :bash:`pylint`
 ---------------------
 
 We are using :bash:`pylint` to check for a variety of possible problems.
-Firecrown is not currently "clean" of all :bash:`pylint` issues, so we are not yet using :bash:`pylint` in the CI testing.
+`pylint` is run in the CI system with the following flags:
+
+.. code:: bash
 
-We are actively working on getting a "clean" report from :bash:`pylint`.
-When this is achieved, we plan to activate :bash:`pylint` in the CI checking.
-This will require that all new code pass :bash:`pylint`'s checks.
+    pylint firecrown
+    pylint --rcfile tests/pylintrc  tests
+    pylint --rcfile firecrown/models/pylintrc firecrown/models
 
 Code formatting
 ---------------
 
 We are using the command-line tool :bash:`black` to auto-format Python code in Firecrown.
 Please make sure to run black on your code before creating any commits.
+
+.. code:: bash
+
+    black firecrown examples tests

docs/dev-notes.rst

@@ -5,15 +5,17 @@ Developer Notes
 .. role:: bash(code)
    :language: bash
 
-To run the current version, one needs to *build* (but not yet install) Firecrown.
-In this directory, run:
+To run the development version, one needs to do an *editable installation* of firecrown.
+This will create an entry in the conda environment that knows about the local code.
+You must have the `:bash:firecrown_developer` conda environment activated before running this installation.
 
-    python3 setup.py build
+Note that we use the :bash:`--no-deps` flag to prevent the installation from accidentally taking in any new dependencies through  :bash:`pip`.
+If the installation fails because of the lack of a dependency, install that dependency using  :bash:`conda` and not  :bash:`pip`.
+If you find a dependency on a package not available through :bash:`conda` please file an issue on the issue tracker.
 
-This will put modules into the subdirectory :bash:`build/lib`.
-Set the environment variable :bash:`FIRECROWN_DIR` to the full path to that directory.
-Set the environment variable :bash:`FIRECROWN_EXAMPLES_DIR` to be the full path to the :bash:`examples` subdirectory.
-
-These environment variables are needed by the example :bash:`ini` files.
-The directory :bash:`$FIRECROWN_DIR` should also be on :bash:`PYTHONPATH`.
+.. code:: bash
+
+    python -m pip install --no-deps --editable ${PWD}
 
+Some of the examples and tests depend on the environment variables :bash:`FIRECROWN_DIR` and :bash:`CSL_DIR`.
+These environment variables are defined for you every time you activate the conda environment.