From e407967518a4d0116e45cb1c387305f40dde5df0 Mon Sep 17 00:00:00 2001
From: benjaminastrand <benjamin01.astrand@gmail.com>
Date: Mon, 25 Nov 2024 14:35:08 +0100
Subject: [PATCH] Move init seed from compute package to separate script

---
 examples/async-clients/README.rst   |  6 ++-
 examples/async-clients/init_seed.py | 65 +++++++++++++++++++++++++++++
 2 files changed, 69 insertions(+), 2 deletions(-)
 create mode 100644 examples/async-clients/init_seed.py

diff --git a/examples/async-clients/README.rst b/examples/async-clients/README.rst
index 45f4a64a8..35c84d04b 100644
--- a/examples/async-clients/README.rst
+++ b/examples/async-clients/README.rst
@@ -38,15 +38,17 @@ Standing in the folder fedn/examples/async-clients
 
     pip install -r requirements.txt
 
-Create the compute package and seed model:
+Create the compute package:
 
 .. code-block::
 
     fedn package create --path client
 
+Create the seed model:
+
 .. code-block::
 
-    fedn run build --path client
+    python init_seed.py seed.npz
 
 
 You will now have a file 'seed.npz' in the directory.
diff --git a/examples/async-clients/init_seed.py b/examples/async-clients/init_seed.py
new file mode 100644
index 000000000..ec20651d7
--- /dev/null
+++ b/examples/async-clients/init_seed.py
@@ -0,0 +1,65 @@
+import sys
+
+import numpy as np
+from sklearn.datasets import make_classification
+from sklearn.model_selection import train_test_split
+from sklearn.neural_network import MLPClassifier
+
+from fedn.utils.helpers.helpers import get_helper
+
+HELPER_MODULE = "numpyhelper"
+
+
+def make_data(n_min=50, n_max=100):
+    """Generate / simulate a random number n data points.
+
+    n will fall in the interval (n_min, n_max)
+
+    """
+    n_samples = 100000
+    X, y = make_classification(n_samples=n_samples, n_features=4, n_informative=4, n_redundant=0, random_state=42)
+    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
+
+    n = np.random.randint(n_min, n_max, 1)[0]
+    ind = np.random.choice(len(X_train), n)
+    X_train = X_train[ind, :]
+    y_train = y_train[ind]
+    return X_train, y_train, X_test, y_test
+
+
+def compile_model(max_iter=1):
+    clf = MLPClassifier(max_iter=max_iter)
+    # This is needed to initialize some state variables needed to make predictions
+    # We will overwrite weights and biases during FL training
+    X_train, y_train, _, _ = make_data()
+    clf.fit(X_train, y_train)
+    return clf
+
+
+def save_parameters(model, out_path):
+    """Save model to disk.
+
+    :param model: The model to save.
+    :type model: torch.nn.Module
+    :param out_path: The path to save to.
+    :type out_path: str
+    """
+    helper = get_helper(HELPER_MODULE)
+    parameters = model.coefs_ + model.intercepts_
+
+    helper.save(parameters, out_path)
+
+
+def init_seed(out_path="seed.npz"):
+    """Initialize seed model.
+
+    :param out_path: The path to save the seed model to.
+    :type out_path: str
+    """
+    # Init and save
+    model = compile_model()
+    save_parameters(model, out_path)
+
+
+if __name__ == "__main__":
+    init_seed(sys.argv[1])