Update installation instructions

README.md

@@ -12,43 +12,67 @@
 
 ## :gear:  Install
 
-We recommend installing `graphnet` in a separate environment, e.g. using Anaconda (see details on installation [here](https://www.anaconda.com/products/individual)) or python virtual environment. This requires specifying a supported python version (see above) and ensuring that the C++ compilers (gcc) are up to date. 
+We recommend installing `graphnet` in a separate environment, e.g. using python virtual environment or Anaconda (see details on installation [here](https://www.anaconda.com/products/individual)). Below we prove installation instructions for different setups.
 
-### Installing stand-alone
+<details>
+<summary><b>Installing with IceTray</b></summary>
+<blockquote>
 
-If you don't need to interface with [IceTray](https://github.com/icecube/icetray/) (e.g., for reading data from I3 files or running inference on these), the following commands should provide a fast way to get up and running on most UNIX systems:
+You may want `graphnet` to be able to interface with IceTray, e.g., when converting I3 files to an intermediate file format for training GNN models (e.g., SQLite or parquet),[^1] or when running GNN inference as part of an IceTray chain. In these cases, you need to install `graphnet` in a python runtime that has IceTray installed.
+
+To achieve this, we recommend running the following commands in a clean bash shell:
 ```bash
-$ git clone git@github.com:<your-username>/graphnet.git
-$ cd graphnet
-$ conda create --name graphnet python=3.8 gcc_linux-64 gxx_linux-64 libgcc -y  # Optional
-$ conda activate graphnet  # Optional
-(graphnet) $ pip install -r requirements/torch_[gpu/cpu].txt -e .[develop,torch]
+$ eval `/cvmfs/icecube.opensciencegrid.org/py3-v4.1.0/setup.sh`
+$ /cvmfs/icecube.opensciencegrid.org/py3-v4.1.0/RHEL_7_x86_64/metaprojects/combo/stable/env-shell.sh
 ```
-If you have an old system version of GCC installed (`< 4.9`), you should add `export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$HOME/anaconda3/lib/` to your `.bashrc` script or similar.
+Optionally, you can alias these commands or save them as a bash script for convenience, as you will have to run these commands every time you want to use IceTray (with `graphnet`) in a clean shell.
 
-This should allow you to e.g. run the scripts in [examples/](./examples/) out of the box. Here, we have installed recent C++ compilers using conda (`gcc_linux-64 gxx_linux-64 libgcc`), but if your system already have recent versions (`$gcc --version` should be > 5, at least) you should be able to omit these from the setup.
+With the IceTray environment active, you can now install `graphnet`, either at a user level or in a python virtual environment. You can either install a light-weight version of `graphnet` without the `torch` extras, i.e., without the machine learning packages (pytorch and pytorch-geometric); this is useful when you just want to convert data from I3 files to, e.g., SQLite, and won't be running inference on I3 files later on. In this case, you don't need to specify a requirements file. If you want torch, you do.
 
-### Installing with IceTray
+<details>
+<summary><b>Install <i>without</i> torch</b></summary>
 
-In some instances, you might want `graphnet` to be able to interface with IceTray, e.g., when converting I3 files to an intermediate file format for training GNN models(e.g., SQLite or parquet), as shown in the [examples/convert_i3_to_sqlite.py](examples/convert_i3_to_sqlite.py) script, or when running GNN inference as part of an IceTray chain. In these cases, you need to install `graphnet` in a python runtime that has IceTray installed.
+```bash
+$ pip install --user -e .[develop]  # Without torch, i.e. only for file conversion
+```
 
-To achieve this, we recommend running the following commands in a clean bash shell:
+</details>
+
+<details>
+<summary><b>Install <i>with</i> torch</b></summary>
 
 ```bash
-$ eval `/cvmfs/icecube.opensciencegrid.org/py3-v4.1.0/setup.sh`
-$ /cvmfs/icecube.opensciencegrid.org/py3-v4.1.0/RHEL_7_x86_64/metaprojects/combo/stable/env-shell.sh
+$ pip install --user -r requirements/torch_cpu.txt -e .[develop,torch]  # CPU-only torch
+$ pip install --user -r requirements/torch_gpu.txt -e .[develop,torch]  # GPU support
 ```
-Optionally, you can alias these commands or save them as a bash script for convenience, as you will have to run these commands every time you want to use IceTray (with `graphnet`) in a clean shell.
 
-With the IceTray environment active, you can now install `graphnet` at a user level. In the example below, we are installing a light-weight version of `graphnet` without the `torch` extras, i.e., without the machine learning packages (pytorch and pytorch-geometric). This is useful when you just want to convert data from I3 files to, e.g., SQLite, and won't be running inference on I3 files later on. In this case, you don't need to specify a requirements file, compared to the example below.
+</details>
+
+This should allow you to run the I3 conversion scripts in [examples/](./examples/) with your preferred I3 files.
+
+</blockquote>
+</details>
+
+<details>
+<summary><b>Installing stand-alone</b></summary>
+<blockquote>
+
+If you don't need to interface with [IceTray](https://github.com/icecube/icetray/) (e.g., for reading data from I3 files or running inference on these), the following commands should provide a fast way to get up and running on most UNIX systems:
 ```bash
-$ conda create --name graphnet_icetray  # Optional
-$ conda activate graphnet_icetray  # Optional
-$ pip install --user -e .[develop]
+$ git clone git@github.com:<your-username>/graphnet.git
+$ cd graphnet
+$ conda create --name graphnet python=3.8 gcc_linux-64 gxx_linux-64 libgcc -y  # Optional
+$ conda activate graphnet  # Optional
+(graphnet) $ pip install -r requirements/torch_cpu.txt -e .[develop,torch]  # CPU-only torch
+(graphnet) $ pip install -r requirements/torch_gpu.txt -e .[develop,torch]  # GPU support
 ```
-Alternatively, you can also install the torch dependencies just as above, remembering the `--user` flag.
+This should allow you to e.g. run the scripts in [examples/](./examples/) out of the box.
+
+A stand-alone installation requires specifying a supported python version (see above) and ensuring that the C++ compilers (gcc) are up to date. Here, we have installed recent C++ compilers using conda (`gcc_linux-64 gxx_linux-64 libgcc`), but if your system already have recent versions (`$gcc --version` should be > 5, at least) you should be able to omit these from the setup.
+If you have an older system version of GCC than this installed and you install newer compiler though the above command, you should add `export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$HOME/anaconda3/lib/` to your `.bashrc` script or similar.
 
-This should allow you to run the [examples/convert_i3_to_sqlite.py](examples/convert_i3_to_sqlite.py) script with your preferred I3 files.
+</blockquote>
+</details>
 
 ## :handshake:  Contributing
 
@@ -75,3 +99,5 @@ $ wandb online
 ```
 
 The [examples/train_model.py](examples/train_model.py) script shows how to train a model and log the results to W&B.
+
+[^1]: Examples of this are shown in the [examples/convert_i3_to_sqlite.py](examples/convert_i3_to_sqlite.py) and [examples/convert_i3_to_parquet.py](examples/convert_i3_to_parquet.py) scripts