Bump black from 22.12.0 to 24.3.0

.github/workflows/publish.yml

@@ -0,0 +1,27 @@
+name: publish
+
+on:
+  push:
+    branches:
+      - main
+
+jobs:
+  publish:
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v2
+
+    - name: Install Poetry
+      run: |
+        python -m pip install --upgrade pip
+        pip install poetry
+        poetry config virtualenvs.in-project false
+        poetry config virtualenvs.path ~/.virtualenvs
+
+    - name: Build and publish to pypi
+      env:
+        PYPI_TOKEN: ${{ secrets.PYPI_TOKEN }}
+      run: |
+        poetry config pypi-token.pypi $PYPI_TOKEN
+        poetry publish --build

.gitignore

@@ -1,14 +1,16 @@
-./data
-.ipynb_checkpoints/
+# python/IDE articfacts
 .idea/
+.ipynb_checkpoints/
+.DS_Store
 __pycache__/
-Deep_Learning_Prediction_Intervals/
-algorithm_implement.py
-exp_dir/
-data/
-!./tests/data/Synthetic Site
-nsrd_auth.json
+coverage/
+dist/
+
+# notebook articfacts
+notebooks/examples/exp_dir/
+notebooks/examples/Synthetic Site/
+notebooks/examples/Training_history.csv
+tests/data/Synthetic Site/Synthetic Site Electricity Main Total Power_external_test.h5
+tests/fixtures/alfa_exp_dir/events.out.tfevents.*
+tests/fixtures/alfa_exp_dir/output.out
 Training_history.csv
-.DS_Store
-.coverage
-dist

README.md

@@ -1,17 +1,16 @@
 Wattile
-====
+=
 
-Deep learning algorithms for probabilistic forecasting of time series building performance data.
+Probabilistic Deep Learning-based Forecasting of Building Energy ConsumptionProbabilistic Deep Learning-based Forecasting of Building Energy Consumption.
 
-Set Up
-----
+# Set Up
 
 Requirements:
 - [Python](https://www.python.org/downloads/) >= 3.7.0
 - [Poetry](https://python-poetry.org/docs/#installation)
 - [Microsoft Visual C++ 14.0](https://visualstudio.microsoft.com/visual-cpp-build-tools/), if you're on Windows
 
-If using conda for python management, use `environment.yml` to make an environment and configure poetry to not create a vitural enviroment.
+If using conda for python management, use `environment.yml` to make an environment and configure poetry to not create a vitural enviroment (else, skip these commands).
 ```
 conda env create -f environment.yml
 poetry config virtualenvs.create false
@@ -29,35 +28,27 @@ poetry run pre-commit install
 
 Quick Start
 ----
+Please see [our example notebooks](./notebooks/examples) to see how to run Wattile out of the box.
 
-Wattile has two main functions:
-- `create_input_dataframe`, which creates a dataframe for model input from the configs file and raw data.
-- `run_model` which runs either training, validation, or prediction on input dataframes according to the configs.
-
-Docs for the configs [here](./docs/Configs.md).
-
-Docs for the format of the raw data is [here](./docs/Data_configs.md).
-
-Docs for the format of the output is [here](./docs/Output.md).
-
+In short, one must prep their data, create the model, and send the data through.
 ```py
-import json 
-
-from wattile.entry_point import create_input_dataframe, run_model
-
-with open("wattile/configs.json", "r") as f:
-    configs = json.load(f)
-
-init_logging(configs["exp_dir"])  
-train_df, val_df = create_input_dataframe(configs)
-run_model(configs, train_df, val_df)
+train_df, val_df = prep_for_rnn(configs, data)
+model = ModelFactory.create_model(configs)
+model.train(train_df, val_df)
 ```
 
 After running, you may use tensordboard on the results.
 
 ```
 tensorboard --logdir=<study directory>
 ```
+Wattile is highly configurable.
+
+Docs for the configs [here](./docs/Configs.md).
+
+Docs for the format of the raw data is [here](./docs/Data_configs.md).
+
+Docs for the format of the output is [here](./docs/Output.md).
 
 Available Models
 ----
@@ -78,6 +69,16 @@ Available Models
 * Vanilla and LSTM variants available
 * Supports future time predictions with constant spacing or variable spacing 
 
+
+### [Alfa Ensemble](./wattile/models/alfa_ensemble_model.py)
+
+* Probabilistic forecasting 
+* In one training session, predict:
+    * Number of future times: *T*
+    * Number of quantiles per time: *Q*
+* Vanilla and LSTM variants available
+* Supports future time predictions with constant spacing or variable spacing 
+
 Development
 ----
 

notebooks/examples/example_batch_training.ipynb

@@ -11,7 +11,7 @@
     "import os\n",
     "from pathlib import Path\n",
     "\n",
-    "from wattile.buildings_processing import prep_for_rnn\n",
+    "from wattile.data_processing import prep_for_rnn\n",
     "from wattile.data_reading import read_dataset_from_file\n",
     "from wattile.entry_point import init_logging\n",
     "from wattile.models import ModelFactory\n",
@@ -42,19 +42,6 @@
     "]"
    ]
   },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "# read configs file for training\n",
-    "with open(\n",
-    "    PROJECT_DIRECTORY / \"wattile\" / \"configs\" / \"configs.json\", \"r\"\n",
-    ") as f:\n",
-    "    configs = json.load(f)"
-   ]
-  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -73,6 +60,19 @@
     "    data_dir = entry[\"dir_data\"]\n",
     "\n",
     "    for target in entry[\"targets\"]:\n",
+    "        \n",
+    "        \"\"\"\n",
+    "        read configs file\n",
+    "        \"\"\"\n",
+    "        with open(\n",
+    "            PROJECT_DIRECTORY / \"wattile\" / \"configs\" / \"configs.json\", \"r\"\n",
+    "        ) as f:\n",
+    "            configs = json.load(f)\n",
+    "            \n",
+    "        \"\"\"\n",
+    "        read configs file\n",
+    "        \"\"\"\n",
+    "        # update configs here only if necessary\n",
     "\n",
     "        \"\"\"\n",
     "        model training\n",
@@ -85,7 +85,7 @@
     "        configs[\"data_input\"][\"data_dir\"] = str(PROJECT_DIRECTORY / data_dir)\n",
     "        configs[\"data_input\"][\"data_config\"] = \"{} Config.json\".format(building)\n",
     "        configs[\"data_input\"][\"target_var\"] = target\n",
-    "        exp_dir = PROJECT_DIRECTORY / building / target\n",
+    "        exp_dir = PROJECT_DIRECTORY / \"notebooks\" / \"examples\" / building / target\n",
     "        configs[\"data_output\"][\"exp_dir\"] = str(exp_dir)\n",
     "\n",
     "        # create results folder\n",