From be88c655a2d03d55229169598fc2011536540d40 Mon Sep 17 00:00:00 2001 From: Nora Loose Date: Mon, 21 Aug 2023 20:17:14 -0600 Subject: [PATCH 1/3] Fix link to contributor guide --- README.md | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/README.md b/README.md index a31296a..19de4f8 100644 --- a/README.md +++ b/README.md @@ -16,16 +16,16 @@ We implement a range of deterministic and stochastic regression baselines to hig ## Getting Started -* [Quick Start](https://leap-stc.github.io/ClimSim/quickstart.html) +* [Quickstart](https://leap-stc.github.io/ClimSim/quickstart.html) * [Dataset Information](https://leap-stc.github.io/ClimSim/dataset.html) -* [Code installation](https://leap-stc.github.io/ClimSim/installation.html) +* [Code Installation](https://leap-stc.github.io/ClimSim/installation.html) * [Baseline Models](https://leap-stc.github.io/ClimSim/models.html) * [Evaluation](https://leap-stc.github.io/ClimSim/evaluating.html) ## Demo notebooks -* [Multi-layer Perceptron (MLP) Example](https://leap-stc.github.io/ClimSim/demo_notebooks/mlp_example.html) +* [Multi-Layer Perceptron (MLP) Example](https://leap-stc.github.io/ClimSim/demo_notebooks/mlp_example.html) * [Convolutional Neural Network (CNN) Example](https://leap-stc.github.io/ClimSim/demo_notebooks/cnn_example.html) -* [Water conservation example](https://leap-stc.github.io/ClimSim/demo_notebooks/water_conservation.html) +* [Water Conservation Example](https://leap-stc.github.io/ClimSim/demo_notebooks/water_conservation.html) ## References * [ClimSim paper](https://arxiv.org/abs/2306.08754) @@ -33,4 +33,4 @@ We implement a range of deterministic and stochastic regression baselines to hig [![Alt text](https://img.youtube.com/vi/M3Vz0zR1Auc/0.jpg)](https://www.youtube.com/watch?v=M3Vz0zR1Auc) -* [Contributor Guide](https://leap-stc.github.io/ClimSim/demo_notebooks/CONTRIBUTING.html) +* [Contributor Guide](https://leap-stc.github.io/ClimSim/CONTRIBUTING.html) From ee7445070e4a3f363003bbe3180bde566d8242a4 Mon Sep 17 00:00:00 2001 From: Nora Loose Date: Mon, 21 Aug 2023 20:17:32 -0600 Subject: [PATCH 2/3] Formatting of notebook captions --- demo_notebooks/cnn_example.ipynb | 2 +- demo_notebooks/mlp_example.ipynb | 2 +- demo_notebooks/water_conservation.ipynb | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/demo_notebooks/cnn_example.ipynb b/demo_notebooks/cnn_example.ipynb index 303b148..227082f 100644 --- a/demo_notebooks/cnn_example.ipynb +++ b/demo_notebooks/cnn_example.ipynb @@ -4,7 +4,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "# CNN example" + "# Convolutional Neural Network (CNN) Example" ] }, { diff --git a/demo_notebooks/mlp_example.ipynb b/demo_notebooks/mlp_example.ipynb index 7d49a2f..7e95528 100644 --- a/demo_notebooks/mlp_example.ipynb +++ b/demo_notebooks/mlp_example.ipynb @@ -4,7 +4,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "# MLP example" + "# Multi-Layer Perceptron (MLP) Example" ] }, { diff --git a/demo_notebooks/water_conservation.ipynb b/demo_notebooks/water_conservation.ipynb index 30ae294..6dd94ec 100644 --- a/demo_notebooks/water_conservation.ipynb +++ b/demo_notebooks/water_conservation.ipynb @@ -4,7 +4,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "# Water conservation example" + "# Water Conservation Example" ] }, { From 5cb61b9b39182acb99bbc0ac56632f6561c0f303 Mon Sep 17 00:00:00 2001 From: Nora Loose Date: Mon, 21 Aug 2023 20:18:25 -0600 Subject: [PATCH 3/3] Formatting and fixing of links --- website/dataset.md | 4 ++-- website/evaluating.md | 2 +- website/installation.md | 2 +- website/quickstart.md | 8 ++++---- 4 files changed, 8 insertions(+), 8 deletions(-) diff --git a/website/dataset.md b/website/dataset.md index 1fd4806..86f77a7 100644 --- a/website/dataset.md +++ b/website/dataset.md @@ -21,10 +21,10 @@ Data from multi-scale climate model (E3SM-MMF) simulations were saved at 20-minu - 1.9 MB per input file, 1.1 MB per output file Input files are labeled ```E3SM-MMF.mli.YYYY-MM-DD-SSSSS.nc```, where ```YYYY-MM-DD-SSS``` corresponds to the simulation year (```YYYY```), month (``MM``), day of the month (``DD``), and seconds of the day (```SSSSS```), with timesteps being spaced 1,200 seconds (20 minutes) apart. Target files are labeled the same way, except ```mli``` is replaced by ```mlo```. -Scalar variables vary in time and "horizontal" grid ("ncol"), while vertically-resolved variables vary additionally in vertical space ("lev"). For vertically-resolved variables, lower indices of "lev" corresponds to higher levels in the atmosphere. This is because pressure decreases monotonically with altitude. +Scalar variables vary in time and "horizontal" grid (`ncol`), while vertically-resolved variables vary additionally in vertical space (`lev`). For vertically-resolved variables, lower indices of `lev` corresponds to higher levels in the atmosphere. This is because pressure decreases monotonically with altitude. The full list of variables can be found in [Supplementary Information](https://arxiv.org/pdf/2306.08754.pdf), Table 1. -There is also a [Quickstart dataset](https://huggingface.co/datasets/LEAP/subsampled_low_res) that contains subsampled and prenormalized data. This data was used for training, validation, and metrics for the ClimSim paper and can be reproduced using the ```preprocessing/create_npy_data_splits.ipynb``` notebook. +There is also a [**Quickstart dataset**](https://huggingface.co/datasets/LEAP/subsampled_low_res) that contains subsampled and prenormalized data. This data was used for training, validation, and metrics for the ClimSim paper and can be reproduced using the [`preprocessing/create_npy_data_splits.ipynb`](https://github.com/leap-stc/ClimSim/tree/main/preprocessing/create_npy_data_splits.ipynb) notebook. diff --git a/website/evaluating.md b/website/evaluating.md index af807db..8af56e6 100644 --- a/website/evaluating.md +++ b/website/evaluating.md @@ -6,7 +6,7 @@ Four different evaluation metrics were calculated: 3. Root mean squared error (RMSE) 4. Continuous ranked probability score (CRPS) -Evaluation and comparison of the different baseline models are found in the ```[evaluation/](https://github.com/leap-stc/ClimSim/tree/main/evaluation)``` folder on GitHub. All variables are converted to a common energy unit (i.e., W/m²) for scoring. The scoring is done using the functions in ```[climsim_utils/data_utils.py](https://github.com/leap-stc/ClimSim/tree/main/climsim_utils)```. +Evaluation and comparison of the different baseline models are found in the [```evaluation/```](https://github.com/leap-stc/ClimSim/tree/main/evaluation) folder on GitHub. All variables are converted to a common energy unit (i.e., W/m²) for scoring. The scoring is done using the functions in [`climsim_utils/data_utils.py`](https://github.com/leap-stc/ClimSim/tree/main/climsim_utils). [This notebook](./evaluation/main_figure_generation.ipynb) calculates and plots MAE, R², RMSE, and CRPS scores for each baseline model. The separate R² for *longitudinally-averaged* and time-averaged 3D variables is found in [this notebook](./evaluation/plot_R2_analysis.ipynb). diff --git a/website/installation.md b/website/installation.md index 4690c1a..097f16a 100644 --- a/website/installation.md +++ b/website/installation.md @@ -1,4 +1,4 @@ -# Installation & setup +# Code Installation For preprocessing and evaluation, please install the `climsim_utils` python tools, by running the following code from the root of this repo: diff --git a/website/quickstart.md b/website/quickstart.md index 7ac9ece..e61431b 100644 --- a/website/quickstart.md +++ b/website/quickstart.md @@ -2,13 +2,13 @@ **Step 1** -The first step is to download the subsampled low-resolution real-geography version of the data [here](https://huggingface.co/datasets/LEAP/subsampled_low_res/tree/main). This contains subsampled and prenormalized data that was used for training, validation, and metrics for the ClimSim paper. It can be reproduced with the full version of the [dataset](https://huggingface.co/datasets/LEAP/ClimSim_low-res) using the [preprocessing/create_npy_data_splits.ipynb](https://github.com/leap-stc/ClimSim/blob/main/preprocessing/create_npy_data_splits.ipynb) notebook. +The first step is to download the subsampled low-resolution real-geography version of the data [here](https://huggingface.co/datasets/LEAP/subsampled_low_res/tree/main). This contains subsampled and prenormalized data that was used for training, validation, and metrics for the ClimSim paper. It can be reproduced with the full version of the [dataset](https://huggingface.co/datasets/LEAP/ClimSim_low-res) using the [`preprocessing/create_npy_data_splits.ipynb`](https://github.com/leap-stc/ClimSim/blob/main/preprocessing/create_npy_data_splits.ipynb) notebook. Training data corresponds to **train_input.npy** and **train_target.npy**. Validation data corresponds to **val_input.npy** and **val_target.npy**. Scoring data (which can be treated as a test set) corresponds to **scoring_input.npy** and **scoring_target.npy**. We have an additional held-out test set that we will use for an upcoming online competition. Keep an eye out! 😉 **Step 2** -Install the `climsim_utils` python tools, by running the following code from the root of this repo: +Install the `climsim_utils` python tools, by running the following code from the root of the [GitHub repo](https://github.com/leap-stc/ClimSim/tree/main): ``` pip install . @@ -22,9 +22,9 @@ pip install . --no-deps **Step 3** -Train your model on the training data and validate using the validation data. If you wish to use something like a CNN, you will probably want to separate the variables into channels and broadcast scalars into vectors of the same dimension as vertically-resolved variables. Methods to do this can be found in the [climsim_utils/data_utils.py](https://github.com/leap-stc/ClimSim/blob/main/climsim_utils/data_utils.py) script. +Train your model on the training data and validate using the validation data. If you wish to use something like a CNN, you will probably want to separate the variables into channels and broadcast scalars into vectors of the same dimension as vertically-resolved variables. Methods to do this can be found in the [`climsim_utils/data_utils.py`](https://github.com/leap-stc/ClimSim/blob/main/climsim_utils/data_utils.py) script. **Step 4** -Evaluation time! Use the [evaluation/main_figure_generation.ipynb](https://github.com/leap-stc/ClimSim/blob/main/evaluation/main_figure_generation.ipynb) notebook to see how your model does! Use the **calc_MAE**, **calc_RMSE**, and **calc_R2** methods in the [climsim_utils/data_utils.py](https://github.com/leap-stc/ClimSim/blob/main/climsim_utils/data_utils.py) script to see how your model does on point estimates and use the calc_CRPS method to check how well-calibrated your model is if it's stochastic. 😊 +Evaluation time! Use the [`evaluation/main_figure_generation.ipynb`](https://github.com/leap-stc/ClimSim/blob/main/evaluation/main_figure_generation.ipynb) notebook to see how your model does! Use the **calc_MAE**, **calc_RMSE**, and **calc_R2** methods in the [`climsim_utils/data_utils.py`](https://github.com/leap-stc/ClimSim/blob/main/climsim_utils/data_utils.py) script to see how your model does on point estimates and use the `calc_CRPS` method to check how well-calibrated your model is if it's stochastic. 😊