From f7fba93d517f4a10948208c9290c8856b9c2615b Mon Sep 17 00:00:00 2001
From: Carl Schmitz <carl.schmitz@connect.qut.edu.au>
Date: Tue, 25 Feb 2025 08:51:10 +1000
Subject: [PATCH] Added info about loading/importing model

---
 CRISPR_DeepEnsemble.ipynb | 486 ++++++++++++++++++++++++++++++++++++--
 1 file changed, 472 insertions(+), 14 deletions(-)

diff --git a/CRISPR_DeepEnsemble.ipynb b/CRISPR_DeepEnsemble.ipynb
index c481f83..ab222b2 100644
--- a/CRISPR_DeepEnsemble.ipynb
+++ b/CRISPR_DeepEnsemble.ipynb
@@ -9,15 +9,13 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [],
    "source": [
     "# Import and set defaults\n",
     "import CRISPR_DeepEnsemble\n",
     "import torch as t\n",
-    "import numpy as np\n",
-    "import pandas as pd\n",
     "from scipy.stats import spearmanr, pearsonr \n",
     "\n",
     "t.manual_seed(123)\n",
@@ -39,7 +37,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -55,7 +53,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -88,9 +86,409 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 4,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/carl/.local/lib/python3.12/site-packages/torch/nn/modules/lazy.py:181: UserWarning: Lazy modules are a new feature under heavy development so changes to the API or functionality can happen at any moment.\n",
+      "  warnings.warn('Lazy modules are a new feature under heavy development '\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Training Model 1/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "  0%|          | 0/50 [00:00<?, ?it/s]/home/carl/.local/lib/python3.12/site-packages/torch/autograd/graph.py:744: UserWarning: CUDA initialization: The NVIDIA driver on your system is too old (found version 11060). Please update your GPU driver by downloading and installing a new version from the URL: http://www.nvidia.com/Download/index.aspx Alternatively, go to: https://pytorch.org to install a PyTorch version that has been compiled with your version of the CUDA driver. (Triggered internally at ../c10/cuda/CUDAFunctions.cpp:108.)\n",
+      "  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass\n",
+      "100%|██████████| 50/50 [04:08<00:00,  4.97s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.829\n",
+      "Training Model 2/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:07<00:00,  4.95s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.802\n",
+      "Training Model 3/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:13<00:00,  5.08s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.803\n",
+      "Training Model 4/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:11<00:00,  5.04s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.803\n",
+      "Training Model 5/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:09<00:00,  4.98s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.864\n",
+      "Training Model 6/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:11<00:00,  5.03s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.761\n",
+      "Training Model 7/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:25<00:00,  5.31s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.827\n",
+      "Training Model 8/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:25<00:00,  5.31s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.877\n",
+      "Training Model 9/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:23<00:00,  5.27s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.775\n",
+      "Training Model 10/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:21<00:00,  5.23s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.790\n",
+      "Training Model 11/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:13<00:00,  5.08s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.768\n",
+      "Training Model 12/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:26<00:00,  5.33s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.765\n",
+      "Training Model 13/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [08:38<00:00, 10.38s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.884\n",
+      "Training Model 14/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:32<00:00,  5.46s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.819\n",
+      "Training Model 15/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:16<00:00,  5.14s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.792\n",
+      "Training Model 16/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:18<00:00,  5.17s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.792\n",
+      "Training Model 17/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:07<00:00,  4.95s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.877\n",
+      "Training Model 18/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:11<00:00,  5.03s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.866\n",
+      "Training Model 19/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:21<00:00,  5.23s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.833\n",
+      "Training Model 20/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:16<00:00,  5.12s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.788\n",
+      "Training Model 21/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:16<00:00,  5.13s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.859\n",
+      "Training Model 22/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:12<00:00,  5.04s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.810\n",
+      "Training Model 23/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:10<00:00,  5.01s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.815\n",
+      "Training Model 24/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:14<00:00,  5.10s/it]\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.776\n",
+      "Training Model 25/25 for 50 epochs...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 50/50 [04:09<00:00,  4.99s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final Epoch Training Loss: -0.755\n",
+      "Training Complete!\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\n"
+     ]
+    }
+   ],
    "source": [
     "# Training a model, simply create a new RegressionDeepEnsemble object and call its `train_ensemble` method.\n",
     "myEnsemble = CRISPR_DeepEnsemble.RegressionDeepEnsemble(BaseNet=CRISPR_DeepEnsemble.CRISPRnet, dataset=myData,\n",
@@ -101,9 +499,20 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 5,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
    "source": [
     "# Example uncertainty plot\n",
     "L,U,IQR = myEnsemble.uncertainty_bounds(inputs =(S_test, F_test), n_samples=1000,\n",
@@ -124,21 +533,70 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 6,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Spearman: 0.843, Pearson: 0.839\n"
+     ]
+    }
+   ],
    "source": [
     "# Compute Spearman correlation between predictions and true answers in test set\n",
     "vals = (pred.numpy(), y_test)\n",
     "print(f\"Spearman: {round(spearmanr(*vals)[0],3)}, Pearson: {round(pearsonr(*vals)[0],3)}\")"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Save model\n",
+    "myEnsemble.save('CRISPR_DeepEnsemble')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Spearman: 0.843, Pearson: 0.839\n"
+     ]
+    }
+   ],
+   "source": [
+    "# Load model\n",
+    "# NOTE: If you are getting import errors when loading the model uncomment the two lines below\n",
+    "# import sys\n",
+    "# sys.modules['CRISPR_DeepEnsemble'] = CRISPR_DeepEnsemble\n",
+    "\n",
+    "newEnsemble = CRISPR_DeepEnsemble.RegressionDeepEnsemble(load_from='CRISPR_DeepEnsemble.zip')\n",
+    "\n",
+    "# Predict values\n",
+    "L,U,IQR = newEnsemble.uncertainty_bounds(inputs =(S_test, F_test), n_samples=1000,\n",
+    "                                    lower=0.01, upper=0.99) \n",
+    "loaded_pred = newEnsemble.predict(inputs = (S_test, F_test))\n",
+    "\n",
+    "# Compute Spearman correlation between predictions and true answers in test set\n",
+    "vals = (loaded_pred.numpy(), y_test)\n",
+    "print(f\"Spearman: {round(spearmanr(*vals)[0],3)}, Pearson: {round(pearsonr(*vals)[0],3)}\")"
+   ]
   }
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "jupyterHub",
+   "display_name": "Python 3",
    "language": "python",
-   "name": "jupyterhub"
+   "name": "python3"
   },
   "language_info": {
    "codemirror_mode": {
@@ -150,7 +608,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.8.10"
+   "version": "3.12.1"
   }
  },
  "nbformat": 4,