notebook

README.md

@@ -1,5 +1,25 @@
-The Portiloop
+# The Portiloop
 
 ![Prototype](https://github.com/nicolasvalenchon/Portiloop/blob/main/images/photo_portiloop.jpg)
 
-Your training curves can be visualized in the Portiloop [wandb project](https://wandb.ai/portiloop).
+Your training curves can be visualized in the Portiloop [wandb project](https://wandb.ai/portiloop).
+
+## Quick start guide
+
+- clone the repo
+- cd to the root of the repo where `setup.py` is
+- pip install with the -e option:
+```terminal
+pip install -e .
+```
+- download the datasets and the experiments zip files
+- unzip the datasets file and paste its content under `Portiloop>Software>dataset`
+- unzip the experiments file and paste its content under `Portiloop>Software>experiments`
+
+### Inference / Portiloop simulation:
+The `simulate_Portiloop_1_input_classification.ipynb` notebook enables stimulating the Portiloop system with and perform inference.
+This notebook can be executed with `jupyter notebook`.
+
+### Training:
+We provide the bash scripts examples for `slurm` to train the model on HPC systems.
+Adapt these scripts to your configuration.

Software/plots/.gitignore

@@ -0,0 +1,4 @@
+# Ignore everything in this directory
+*
+# Except this file
+!.gitignore

Software/python/ANN/portiloop_detector_training.py

@@ -68,7 +68,7 @@ def __init__(self, filename, path, window_size, fe, seq_len, seq_stride, list_su
                         if not (self.data[3][idx + self.window_size - 1] < 0  # that are not ending in an unlabeled zone
                                 or idx < self.past_signal_len)]  # and far enough from the beginning to build a sequence up to here
         total_spindles = np.sum(self.data[3] > THRESHOLD)
-        logging.debug(f"nb total of spindles in this dataset : {total_spindles}")
+        logging.debug(f"total number of spindles in this dataset : {total_spindles}")
 
     def __len__(self):
         return len(self.indices)