Merge pull request #35 from gaelforget/v0p1p13b

V0p1p13b

Project.toml

@@ -1,7 +1,7 @@
 name = "ClimateModels"
 uuid = "f6adb021-9183-4f40-84dc-8cea6f651bb0"
 authors = ["gaelforget <[email protected]>"]
-version = "0.1.12"
+version = "0.1.13"
 
 [deps]
 AWS = "fbe9abb3-538b-5e4e-ba9e-bc94f4f92ebc"

README.md

@@ -25,6 +25,12 @@ It also supports e.g. cloud computing workflows that start from previous model o
 
 - [CMIP6 model output](https://gaelforget.github.io/ClimateModels.jl/dev/generated/CMIP6/) via cloud storage 
 
+## Trying Out The Examples
+
+For detail please refer to [the docs](https://gaelforget.github.io/ClimateModels.jl/dev/). All requirements are preinstalled in the following <https://mybinder.org> cloud instance. 
+
+[![Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/gh/gaelforget/ClimateModels.jl/HEAD?urlpath=lab)
+
 ## JuliaCon 2021 Presentation Link
 
 [![Screen Shot 2021-08-31 at 2 25 04 PM](https://user-images.githubusercontent.com/20276764/131556274-48f3df13-0608-4cd0-acf9-c3e29894a32c.png)](https://youtu.be/XR5hKCja0uw)

docs/src/examples.md

@@ -16,21 +16,21 @@ The next sections are examples which broadly fall into two categories.
 
 ## [Running The Examples](@id examples-running)
 
-An example found in the online documentation can be run as follows:
+Any example found in the online documentation can be run as follows:
 
 ```
 git clone https://github.com/gaelforget/ClimateModels.jl
 cd ClimateModels.jl
 
-julia --project=docs/ 'using Pkg; Pkg.instantiate()'
+julia --project=docs/ -e 'using Pkg; Pkg.instantiate()'
 julia --project=docs/ -e 'using Pkg; Pkg.develop(PackageSpec(path=pwd()))'
 
 julia --project=docs/ -e 'include("examples/defaults.jl")'
 ```
 
 The example file name (`defaults.jl` here) can readily be replaced with another one from `examples/` to run a different example. However, running examples which rely on a fortran compiler (`gfortran`) and / or netcdf libraries (`libnetcdf-dev`,`libnetcdff-dev`) will require that those have been pre-installed. It should also be noted that some models may only support linux based environments (i.e. linux and macos).
 
-All requirements are preinstalled in the following <https://mybinder.org> cloud instance, where one can just open a terminal window and run the above commands. All examples have been run in this fashion.
+All requirements are preinstalled in the <https://mybinder.org> cloud instance linked below, where one can just open a terminal window and run the above _julia  --project=docs/ ..._ commands directly. The final command should return the _run folder_ path name where outputs will be located.
 
 [![Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/gh/gaelforget/ClimateModels.jl/HEAD?urlpath=lab)
 

docs/src/functionalities.md

@@ -4,7 +4,7 @@ Here we document key functionalities offered in `ClimateModels.jl`
 
 ## Climate Model Interface
 
-The climate model interface is based on a data structure (`ModelConfig`) and a series of methods like `setup` and `launch`. The default assumption is that the model is either _1)_ a `Julia` package to be downloaded from a URL within `setup` using `Pkg`, and run within `launch` using `Pkg.test` or _2)_ a `Julia` function to be called with a `ModelConfig` argument. 
+The climate model interface is based on a data structure (`ModelConfig`) and a series of methods like `setup`, `build`, and `launch`. The default assumption is that the model is either _1)_ a `Julia` package to be downloaded from a URL within `setup` using `Pkg.develop`, and run within `launch` using `Pkg.test` or _2)_ a `Julia` function to be called with a `ModelConfig` argument. 
 
 A key point is that everything can be customized to e.g. `1)` use a custom workflow instead of `Pkg.test` or `2)` use popular models previously written in Fortran or C just as simply. The latter typically involves calling a `build` method to compile the model between `setup` and `launch`.
 

examples/CMIP6.jl

@@ -52,6 +52,7 @@ MC=ModelConfig(model="GlobalAverage",configuration=GlobalAverage,inputs=paramete
 # _Note: this step may take most time, since `launch` is where data is accessed over the internet, and computation takes place.
 
 setup(MC)
+build(MC)
 launch(MC)
 
 # ## Read Output Files

examples/RandomWalker.jl

@@ -32,8 +32,8 @@ end
 
 MC=ModelConfig(model=RandomWalker,inputs=Dict("nSteps" => 1000))
 setup(MC)
+build(MC)
 launch(MC)
-MC
 
 # ## Exercise 
 #

examples/ShallowWaters.jl

@@ -26,11 +26,12 @@ end
 #
 # `ModelConfig` wraps up the model into a data structure, `MC`, which also includes e.g. the online location for the model repository, parameters, and a local folder path used later on.
 
-MC=ModelConfig(model=URL,configuration=SWM,inputs=parameters)
+MC=ModelConfig(model=URL,configuration=SWM,inputs=parameters);
 
 # The `setup` function then calls `Pkg.develop` and sets up the `git` log subfolder.
 
 setup(MC)
+build(MC)
 
 # ## Run Model
 #

examples/defaults.jl

@@ -19,6 +19,7 @@ MC=ModelConfig(model=url)
 #
 
 setup(MC)
+build(MC)
 launch(MC)
 
 # See run folder for workflow output:

src/interface.jl

@@ -195,10 +195,7 @@ true
 compile(x :: AbstractModelConfig) = default_ClimateModelBuild(x)
 
 function default_ClimateModelBuild(x::AbstractModelConfig)
-    @suppress begin
-        isa(x.model,String) ? Pkg.build(x.model) : nothing
-        isa(x.model,Pkg.Types.PackageSpec) ? build_the_pkg(x) : nothing
-    end
+    isa(x.model,Pkg.Types.PackageSpec) ? build_the_pkg(x) : nothing
 end
 
 """