733 lines
114 KiB
Plaintext
733 lines
114 KiB
Plaintext
{
|
|
"cells": [
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "5d904dee",
|
|
"metadata": {},
|
|
"source": [
|
|
"# Community 2: Protein Sequence Classification"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "046905a8",
|
|
"metadata": {},
|
|
"source": [
|
|
"**Disclaimer: This is uploaded from a github user, not the KAN authors. KAN authors did not writer this or proofread this carefully, hence are not responsible for mistakes in this notebook. If you have questions, please consult the github user who uploaded it. Thank you!**"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "da7e6ae8",
|
|
"metadata": {},
|
|
"source": [
|
|
"In this example, we will see how to use KAN in protein sequence classification. We will be using one hot encoding to encode the amino acids."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "8c7e3d97-d0f6-41bc-8799-ad9c4a608a88",
|
|
"metadata": {},
|
|
"source": [
|
|
"#### This is just an example how it can be used for protein sequences. Need to use real data to actually observe the performance."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 1,
|
|
"id": "0a59179d",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"from kan import *\n",
|
|
"import torch\n",
|
|
"import random\n",
|
|
"import numpy as np"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 2,
|
|
"id": "3f0cd8cd-1161-4dd1-bbdc-31efe46f78c3",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# Hyperparameters\n",
|
|
"PROTEIN_WINDOW_SIZE = 5 \n",
|
|
"\n",
|
|
"# define the universe of possible input amino acids, ie. vocab list\n",
|
|
"aa_list = 'ARNDCQEGHILKMFPSTWYVX'"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 3,
|
|
"id": "25e9f373-3755-4d53-8529-dcf4b71acf18",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"def one_hot_encode(protein_sequence):\n",
|
|
" \"\"\"\n",
|
|
" One-hot encodes a protein sequence.\n",
|
|
"\n",
|
|
" Args:\n",
|
|
" protein_sequence (str): The input protein sequence.\n",
|
|
"\n",
|
|
" Returns:\n",
|
|
" numpy.array: The one-hot encoded representation of the protein sequence.\n",
|
|
" \"\"\"\n",
|
|
" # Create a dictionary mapping amino acids to indices\n",
|
|
" aa_to_index = {aa: i for i, aa in enumerate(aa_list)}\n",
|
|
" \n",
|
|
" # Initialize an array of zeros with shape (sequence_length, alphabet_length)\n",
|
|
" encoding = np.zeros((len(protein_sequence), len(aa_list)))\n",
|
|
" \n",
|
|
" # Iterate over the protein sequence and set the corresponding index to 1\n",
|
|
" for i, aa in enumerate(protein_sequence):\n",
|
|
" if aa in aa_to_index:\n",
|
|
" encoding[i, aa_to_index[aa]] = 1\n",
|
|
" else:\n",
|
|
" # If the amino acid is not in the alphabet, set the last index to 1 (unknown)\n",
|
|
" encoding[i, -1] = 1\n",
|
|
" \n",
|
|
" return encoding"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 5,
|
|
"id": "90b53975-dd55-4ae0-816f-a4ed5cce7e23",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"GTKYX 1\n",
|
|
"TTKPP 1\n",
|
|
"AESVY 0\n",
|
|
"MYSFD 0\n",
|
|
"SQKNT 1\n",
|
|
"IDKAC 1\n",
|
|
"AXKTA 1\n",
|
|
"TESDW 0\n",
|
|
"YXSTF 0\n",
|
|
"VTSYF 0\n",
|
|
"HYKYE 1\n",
|
|
"RDSPA 0\n",
|
|
"MDSNK 0\n",
|
|
"SCKFH 1\n",
|
|
"AHKED 1\n",
|
|
"EFKYA 1\n",
|
|
"EPKLR 1\n",
|
|
"GWSRE 0\n",
|
|
"GMSYE 0\n",
|
|
"IPSKD 0\n",
|
|
"NSKQA 1\n",
|
|
"TWKNL 1\n",
|
|
"TCKFF 1\n",
|
|
"HNKSG 1\n",
|
|
"QNSKR 0\n",
|
|
"RVKYC 1\n",
|
|
"TESCP 0\n",
|
|
"SMKXE 1\n",
|
|
"IYSEV 0\n",
|
|
"XQSKD 0\n",
|
|
"VKSYN 0\n",
|
|
"EESGV 0\n",
|
|
"IISMQ 0\n",
|
|
"FLKGE 1\n",
|
|
"VMKGH 1\n",
|
|
"PTKMH 1\n",
|
|
"TLSIQ 0\n",
|
|
"TTSMA 0\n",
|
|
"ATKEE 1\n",
|
|
"MGSFT 0\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"def generate_sample_protein_dataset(num_samples=20, protein_window_size=5):\n",
|
|
" \"\"\"\n",
|
|
" Generate a dataset of protein sequences of length 11, keeping Lysine(K) in the center for label 1 and Serine(S) for label 0. \n",
|
|
"\n",
|
|
" Args:\n",
|
|
" num_samples (int): Number of samples to generate.\n",
|
|
" protein_window_size (int): Length of the protein sequence.\n",
|
|
"\n",
|
|
" Returns:\n",
|
|
" dict: A dictionary containing train_input, test_input, train_label, and test_label.\n",
|
|
" \"\"\"\n",
|
|
" \n",
|
|
" dataset = {'train_input': [], 'test_input': [], 'train_label': [], 'test_label': []}\n",
|
|
" alphabet = 'ARNDCQEGHILKMFPSTWYVX'\n",
|
|
"\n",
|
|
" # Generate half of the samples with label 1 and half with label 0\n",
|
|
" label_sequence = [1] * (num_samples // 2) + [0] * (num_samples // 2)\n",
|
|
" random.shuffle(label_sequence)\n",
|
|
"\n",
|
|
" for label in label_sequence:\n",
|
|
" # Generate a protein sequence with 'K' in the middle for label 1 and 'S' for label 0\n",
|
|
" if label == 1:\n",
|
|
" center_aa = 'K'\n",
|
|
" else:\n",
|
|
" center_aa = 'S'\n",
|
|
" sequence = ''.join(random.choices(alphabet.replace(center_aa, ''), k=protein_window_size//2)) + center_aa + ''.join(random.choices(alphabet.replace(center_aa, ''), k=protein_window_size//2))\n",
|
|
" print(sequence, label)\n",
|
|
" encoded_sequence = one_hot_encode(sequence).flatten()\n",
|
|
"\n",
|
|
" # Split the dataset into train and test (50% each)\n",
|
|
" if len(dataset['train_input']) < num_samples // 2:\n",
|
|
" dataset['train_input'].append(encoded_sequence)\n",
|
|
" dataset['train_label'].append(label)\n",
|
|
" else:\n",
|
|
" dataset['test_input'].append(encoded_sequence)\n",
|
|
" dataset['test_label'].append(label)\n",
|
|
"\n",
|
|
" # Convert lists to tensors\n",
|
|
" dataset['train_input'] = torch.tensor(dataset['train_input'])\n",
|
|
" dataset['test_input'] = torch.tensor(dataset['test_input'])\n",
|
|
" dataset['train_label'] = torch.tensor(dataset['train_label']).view(-1, 1)\n",
|
|
" dataset['test_label'] = torch.tensor(dataset['test_label']).view(-1, 1)\n",
|
|
"\n",
|
|
" return dataset\n",
|
|
"\n",
|
|
"# Generate dataset with 10 samples\n",
|
|
"dataset = generate_sample_protein_dataset(40)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 6,
|
|
"id": "44e5b378-0d30-4886-8d4f-bc8c515a8e95",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"{'train_input': tensor([[0., 0., 0., ..., 0., 0., 1.],\n",
|
|
" [0., 0., 0., ..., 0., 0., 0.],\n",
|
|
" [1., 0., 0., ..., 1., 0., 0.],\n",
|
|
" ...,\n",
|
|
" [0., 0., 0., ..., 0., 0., 0.],\n",
|
|
" [0., 0., 0., ..., 0., 0., 0.],\n",
|
|
" [0., 0., 0., ..., 0., 0., 0.]], dtype=torch.float64), 'test_input': tensor([[0., 0., 1., ..., 0., 0., 0.],\n",
|
|
" [0., 0., 0., ..., 0., 0., 0.],\n",
|
|
" [0., 0., 0., ..., 0., 0., 0.],\n",
|
|
" ...,\n",
|
|
" [0., 0., 0., ..., 0., 0., 0.],\n",
|
|
" [1., 0., 0., ..., 0., 0., 0.],\n",
|
|
" [0., 0., 0., ..., 0., 0., 0.]], dtype=torch.float64), 'train_label': tensor([[1],\n",
|
|
" [1],\n",
|
|
" [0],\n",
|
|
" [0],\n",
|
|
" [1],\n",
|
|
" [1],\n",
|
|
" [1],\n",
|
|
" [0],\n",
|
|
" [0],\n",
|
|
" [0],\n",
|
|
" [1],\n",
|
|
" [0],\n",
|
|
" [0],\n",
|
|
" [1],\n",
|
|
" [1],\n",
|
|
" [1],\n",
|
|
" [1],\n",
|
|
" [0],\n",
|
|
" [0],\n",
|
|
" [0]]), 'test_label': tensor([[1],\n",
|
|
" [1],\n",
|
|
" [1],\n",
|
|
" [1],\n",
|
|
" [0],\n",
|
|
" [1],\n",
|
|
" [0],\n",
|
|
" [1],\n",
|
|
" [0],\n",
|
|
" [0],\n",
|
|
" [0],\n",
|
|
" [0],\n",
|
|
" [0],\n",
|
|
" [1],\n",
|
|
" [1],\n",
|
|
" [1],\n",
|
|
" [0],\n",
|
|
" [0],\n",
|
|
" [1],\n",
|
|
" [0]])}\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"print(dataset)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 9,
|
|
"id": "fe465888-e3f2-4f06-bfc2-b93ff17eab63",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# define model\n",
|
|
"# create a KAN: 105 inputs, 2D output, and 3 hidden neurons. k=2, 3 grid intervals (grid=3).\n",
|
|
"# considering window size: 5, 5 times 21(vocab size), input-> 21 * 5\n",
|
|
"\n",
|
|
"model = KAN(width=[105,3,2], grid=3, k=2)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 10,
|
|
"id": "e9aa3305-f6da-438d-bf86-36eb12fa4e5f",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"train loss: 1.04e-03 | test loss: 2.33e-01 | reg: 6.38e+01 : 100%|████| 5/5 [00:15<00:00, 3.00s/it]\n"
|
|
]
|
|
},
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"(1.0, 0.949999988079071)"
|
|
]
|
|
},
|
|
"execution_count": 10,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"def train_acc():\n",
|
|
" return torch.mean((torch.round(model(dataset['train_input'])[:,0]) == dataset['train_label'][:,0]).float())\n",
|
|
"\n",
|
|
"def test_acc():\n",
|
|
" return torch.mean((torch.round(model(dataset['test_input'])[:,0]) == dataset['test_label'][:,0]).float())\n",
|
|
"\n",
|
|
"results = model.train(dataset, opt=\"LBFGS\", steps=5, metrics=(train_acc, test_acc));\n",
|
|
"results['train_acc'][-1], results['test_acc'][-1]"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 11,
|
|
"id": "0bcb80ed-e5fa-456f-8910-15c82e4fd6c0",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"fixing (0,0,0) with x^2, r2=0.9999999665312771\n",
|
|
"fixing (0,0,1) with x^2, r2=0.9999979934036755\n",
|
|
"fixing (0,0,2) with x^2, r2=0.9999999622133074\n",
|
|
"fixing (0,1,0) with x^2, r2=0.9999999799949156\n",
|
|
"fixing (0,1,1) with x^2, r2=0.9991883825579457\n",
|
|
"fixing (0,1,2) with x^2, r2=0.9999994895376765\n",
|
|
"fixing (0,2,0) with x^2, r2=0.9999990593107048\n",
|
|
"fixing (0,2,1) with x^2, r2=0.9999996655563207\n",
|
|
"fixing (0,2,2) with x^2, r2=0.999999966951783\n",
|
|
"fixing (0,3,0) with x^2, r2=0.0\n",
|
|
"fixing (0,3,1) with x^2, r2=0.0\n",
|
|
"fixing (0,3,2) with x^2, r2=0.0\n",
|
|
"fixing (0,4,0) with x^2, r2=0.0\n",
|
|
"fixing (0,4,1) with x^2, r2=0.0\n",
|
|
"fixing (0,4,2) with x^2, r2=0.0\n",
|
|
"fixing (0,5,0) with x^2, r2=0.9999998808271742\n",
|
|
"fixing (0,5,1) with x^2, r2=0.9999998953621121\n",
|
|
"fixing (0,5,2) with x^2, r2=0.999999968375537\n",
|
|
"fixing (0,6,0) with x^2, r2=0.9981315108075913\n",
|
|
"fixing (0,6,1) with x^2, r2=0.999999843899342\n",
|
|
"fixing (0,6,2) with x^2, r2=0.9999999589830514\n",
|
|
"fixing (0,7,0) with x^2, r2=0.0\n",
|
|
"fixing (0,7,1) with x^2, r2=0.0\n",
|
|
"fixing (0,7,2) with x^2, r2=0.0\n",
|
|
"fixing (0,8,0) with x^2, r2=0.9999998200480685\n",
|
|
"fixing (0,8,1) with x^2, r2=0.9999999862277233\n",
|
|
"fixing (0,8,2) with x^2, r2=0.9999813684975204\n",
|
|
"fixing (0,9,0) with x^2, r2=0.9999999870502827\n",
|
|
"fixing (0,9,1) with x^2, r2=0.9997068764841773\n",
|
|
"fixing (0,9,2) with x^2, r2=0.9999999768060073\n",
|
|
"fixing (0,10,0) with x^2, r2=0.0\n",
|
|
"fixing (0,10,1) with x^2, r2=0.0\n",
|
|
"fixing (0,10,2) with x^2, r2=0.0\n",
|
|
"fixing (0,11,0) with x^2, r2=0.0\n",
|
|
"fixing (0,11,1) with x^2, r2=0.0\n",
|
|
"fixing (0,11,2) with x^2, r2=0.0\n",
|
|
"fixing (0,12,0) with x^2, r2=0.9999996829291468\n",
|
|
"fixing (0,12,1) with x^2, r2=0.9999747579126426\n",
|
|
"fixing (0,12,2) with x^2, r2=0.999999983307972\n",
|
|
"fixing (0,13,0) with x^2, r2=0.9999999625943928\n",
|
|
"fixing (0,13,1) with x^2, r2=0.9999999376278957\n",
|
|
"fixing (0,13,2) with x^2, r2=0.9999982391574459\n",
|
|
"fixing (0,14,0) with x^2, r2=0.9999999540837675\n",
|
|
"fixing (0,14,1) with x^2, r2=0.999993702906714\n",
|
|
"fixing (0,14,2) with x^2, r2=0.9999996570009488\n",
|
|
"fixing (0,15,0) with x^2, r2=0.999994330617256\n",
|
|
"fixing (0,15,1) with x^2, r2=0.9999996275829637\n",
|
|
"fixing (0,15,2) with x^2, r2=0.9999999847151517\n",
|
|
"fixing (0,16,0) with x^2, r2=0.9999999965050976\n",
|
|
"fixing (0,16,1) with x^2, r2=0.9999999736671104\n",
|
|
"fixing (0,16,2) with x^2, r2=0.9999999930306683\n",
|
|
"fixing (0,17,0) with x^2, r2=0.0\n",
|
|
"fixing (0,17,1) with x^2, r2=0.0\n",
|
|
"fixing (0,17,2) with x^2, r2=0.0\n",
|
|
"fixing (0,18,0) with x^2, r2=0.0\n",
|
|
"fixing (0,18,1) with x^2, r2=0.0\n",
|
|
"fixing (0,18,2) with x^2, r2=0.0\n",
|
|
"fixing (0,19,0) with x^2, r2=0.9999999090971862\n",
|
|
"fixing (0,19,1) with x^2, r2=0.999999811862135\n",
|
|
"fixing (0,19,2) with x^2, r2=0.9999989774097001\n",
|
|
"fixing (0,20,0) with x^2, r2=0.9999998410838922\n",
|
|
"fixing (0,20,1) with x^2, r2=0.999999954524944\n",
|
|
"fixing (0,20,2) with x^2, r2=0.9999995236701958\n",
|
|
"fixing (0,21,0) with x^2, r2=0.0\n",
|
|
"fixing (0,21,1) with x^2, r2=0.0\n",
|
|
"fixing (0,21,2) with x^2, r2=0.0\n",
|
|
"fixing (0,22,0) with x^2, r2=0.0\n",
|
|
"fixing (0,22,1) with x^2, r2=0.0\n",
|
|
"fixing (0,22,2) with x^2, r2=0.0\n",
|
|
"fixing (0,23,0) with x^2, r2=0.9999999953439344\n",
|
|
"fixing (0,23,1) with x^2, r2=0.9999999811625986\n",
|
|
"fixing (0,23,2) with x^2, r2=0.9999999555240675\n",
|
|
"fixing (0,24,0) with x^2, r2=0.0\n",
|
|
"fixing (0,24,1) with x^2, r2=0.0\n",
|
|
"fixing (0,24,2) with x^2, r2=0.0\n",
|
|
"fixing (0,25,0) with x^2, r2=0.9999998811160122\n",
|
|
"fixing (0,25,1) with x^2, r2=0.9999999304599131\n",
|
|
"fixing (0,25,2) with x^2, r2=0.9999998146150727\n",
|
|
"fixing (0,26,0) with x^2, r2=0.9999984806067732\n",
|
|
"fixing (0,26,1) with x^2, r2=0.9999999378197437\n",
|
|
"fixing (0,26,2) with x^2, r2=0.9999994597119173\n",
|
|
"fixing (0,27,0) with x^2, r2=0.9999991631417857\n",
|
|
"fixing (0,27,1) with x^2, r2=0.9999995673636365\n",
|
|
"fixing (0,27,2) with x^2, r2=0.9999999532647686\n",
|
|
"fixing (0,28,0) with x^2, r2=0.9999999703007609\n",
|
|
"fixing (0,28,1) with x^2, r2=0.999999684803164\n",
|
|
"fixing (0,28,2) with x^2, r2=0.9999999512126377\n",
|
|
"fixing (0,29,0) with x^2, r2=0.0\n",
|
|
"fixing (0,29,1) with x^2, r2=0.0\n",
|
|
"fixing (0,29,2) with x^2, r2=0.0\n",
|
|
"fixing (0,30,0) with x^2, r2=0.9999999361143834\n",
|
|
"fixing (0,30,1) with x^2, r2=0.9999999526237395\n",
|
|
"fixing (0,30,2) with x^2, r2=0.9999999758476676\n",
|
|
"fixing (0,31,0) with x^2, r2=0.9999999772937739\n",
|
|
"fixing (0,31,1) with x^2, r2=0.999998823370015\n",
|
|
"fixing (0,31,2) with x^2, r2=0.9999999951682172\n",
|
|
"fixing (0,32,0) with x^2, r2=0.9999998454496639\n",
|
|
"fixing (0,32,1) with x^2, r2=0.9999902771971996\n",
|
|
"fixing (0,32,2) with x^2, r2=0.9993939197671529\n",
|
|
"fixing (0,33,0) with x^2, r2=0.9979543880597602\n",
|
|
"fixing (0,33,1) with x^2, r2=0.9999999733685552\n",
|
|
"fixing (0,33,2) with x^2, r2=0.9999999872961335\n",
|
|
"fixing (0,34,0) with x^2, r2=0.0\n",
|
|
"fixing (0,34,1) with x^2, r2=0.0\n",
|
|
"fixing (0,34,2) with x^2, r2=0.0\n",
|
|
"fixing (0,35,0) with x^2, r2=0.0\n",
|
|
"fixing (0,35,1) with x^2, r2=0.0\n",
|
|
"fixing (0,35,2) with x^2, r2=0.0\n",
|
|
"fixing (0,36,0) with x^2, r2=0.9999997063428989\n",
|
|
"fixing (0,36,1) with x^2, r2=0.9999999499783073\n",
|
|
"fixing (0,36,2) with x^2, r2=0.9999997789665279\n",
|
|
"fixing (0,37,0) with x^2, r2=0.9999999009788131\n",
|
|
"fixing (0,37,1) with x^2, r2=0.9999999715302882\n",
|
|
"fixing (0,37,2) with x^2, r2=0.9999994175010077\n",
|
|
"fixing (0,38,0) with x^2, r2=0.9999998691174623\n",
|
|
"fixing (0,38,1) with x^2, r2=0.9999932563050576\n",
|
|
"fixing (0,38,2) with x^2, r2=0.9999999113693885\n",
|
|
"fixing (0,39,0) with x^2, r2=0.9999998298601666\n",
|
|
"fixing (0,39,1) with x^2, r2=0.9999889526353061\n",
|
|
"fixing (0,39,2) with x^2, r2=0.9999999603098101\n",
|
|
"fixing (0,40,0) with x^2, r2=0.9999941430142316\n",
|
|
"fixing (0,40,1) with x^2, r2=0.9999907490633038\n",
|
|
"fixing (0,40,2) with x^2, r2=0.9999999184598747\n",
|
|
"fixing (0,41,0) with x^2, r2=0.0\n",
|
|
"fixing (0,41,1) with x^2, r2=0.0\n",
|
|
"fixing (0,41,2) with x^2, r2=0.0\n",
|
|
"fixing (0,42,0) with x^2, r2=0.0\n",
|
|
"fixing (0,42,1) with x^2, r2=0.0\n",
|
|
"fixing (0,42,2) with x^2, r2=0.0\n",
|
|
"fixing (0,43,0) with x^2, r2=0.0\n",
|
|
"fixing (0,43,1) with x^2, r2=0.0\n",
|
|
"fixing (0,43,2) with x^2, r2=0.0\n",
|
|
"fixing (0,44,0) with x^2, r2=0.0\n",
|
|
"fixing (0,44,1) with x^2, r2=0.0\n",
|
|
"fixing (0,44,2) with x^2, r2=0.0\n",
|
|
"fixing (0,45,0) with x^2, r2=0.0\n",
|
|
"fixing (0,45,1) with x^2, r2=0.0\n",
|
|
"fixing (0,45,2) with x^2, r2=0.0\n",
|
|
"fixing (0,46,0) with x^2, r2=0.0\n",
|
|
"fixing (0,46,1) with x^2, r2=0.0\n",
|
|
"fixing (0,46,2) with x^2, r2=0.0\n",
|
|
"fixing (0,47,0) with x^2, r2=0.0\n",
|
|
"fixing (0,47,1) with x^2, r2=0.0\n",
|
|
"fixing (0,47,2) with x^2, r2=0.0\n",
|
|
"fixing (0,48,0) with x^2, r2=0.0\n",
|
|
"fixing (0,48,1) with x^2, r2=0.0\n",
|
|
"fixing (0,48,2) with x^2, r2=0.0\n",
|
|
"fixing (0,49,0) with x^2, r2=0.0\n",
|
|
"fixing (0,49,1) with x^2, r2=0.0\n",
|
|
"fixing (0,49,2) with x^2, r2=0.0\n",
|
|
"fixing (0,50,0) with x^2, r2=0.0\n",
|
|
"fixing (0,50,1) with x^2, r2=0.0\n",
|
|
"fixing (0,50,2) with x^2, r2=0.0\n",
|
|
"fixing (0,51,0) with x^2, r2=0.0\n",
|
|
"fixing (0,51,1) with x^2, r2=0.0\n",
|
|
"fixing (0,51,2) with x^2, r2=0.0\n",
|
|
"fixing (0,52,0) with x^2, r2=0.0\n",
|
|
"fixing (0,52,1) with x^2, r2=0.0\n",
|
|
"fixing (0,52,2) with x^2, r2=0.0\n",
|
|
"fixing (0,53,0) with x^2, r2=0.9999999987614517\n",
|
|
"fixing (0,53,1) with x^2, r2=0.9999999995688087\n",
|
|
"fixing (0,53,2) with x^2, r2=0.999999999716506\n",
|
|
"fixing (0,54,0) with x^2, r2=0.0\n",
|
|
"fixing (0,54,1) with x^2, r2=0.0\n",
|
|
"fixing (0,54,2) with x^2, r2=0.0\n",
|
|
"fixing (0,55,0) with x^2, r2=0.0\n",
|
|
"fixing (0,55,1) with x^2, r2=0.0\n",
|
|
"fixing (0,55,2) with x^2, r2=0.0\n",
|
|
"fixing (0,56,0) with x^2, r2=0.0\n",
|
|
"fixing (0,56,1) with x^2, r2=0.0\n",
|
|
"fixing (0,56,2) with x^2, r2=0.0\n",
|
|
"fixing (0,57,0) with x^2, r2=0.9999999977865017\n",
|
|
"fixing (0,57,1) with x^2, r2=0.999999999143338\n",
|
|
"fixing (0,57,2) with x^2, r2=0.9999999998290019\n",
|
|
"fixing (0,58,0) with x^2, r2=0.0\n",
|
|
"fixing (0,58,1) with x^2, r2=0.0\n",
|
|
"fixing (0,58,2) with x^2, r2=0.0\n",
|
|
"fixing (0,59,0) with x^2, r2=0.0\n",
|
|
"fixing (0,59,1) with x^2, r2=0.0\n",
|
|
"fixing (0,59,2) with x^2, r2=0.0\n",
|
|
"fixing (0,60,0) with x^2, r2=0.0\n",
|
|
"fixing (0,60,1) with x^2, r2=0.0\n",
|
|
"fixing (0,60,2) with x^2, r2=0.0\n",
|
|
"fixing (0,61,0) with x^2, r2=0.0\n",
|
|
"fixing (0,61,1) with x^2, r2=0.0\n",
|
|
"fixing (0,61,2) with x^2, r2=0.0\n",
|
|
"fixing (0,62,0) with x^2, r2=0.0\n",
|
|
"fixing (0,62,1) with x^2, r2=0.0\n",
|
|
"fixing (0,62,2) with x^2, r2=0.0\n",
|
|
"fixing (0,63,0) with x^2, r2=0.0\n",
|
|
"fixing (0,63,1) with x^2, r2=0.0\n",
|
|
"fixing (0,63,2) with x^2, r2=0.0\n",
|
|
"fixing (0,64,0) with x^2, r2=0.0\n",
|
|
"fixing (0,64,1) with x^2, r2=0.0\n",
|
|
"fixing (0,64,2) with x^2, r2=0.0\n",
|
|
"fixing (0,65,0) with x^2, r2=0.9999999302979558\n",
|
|
"fixing (0,65,1) with x^2, r2=0.9999902406071391\n",
|
|
"fixing (0,65,2) with x^2, r2=0.9999998684472524\n",
|
|
"fixing (0,66,0) with x^2, r2=0.0\n",
|
|
"fixing (0,66,1) with x^2, r2=0.0\n",
|
|
"fixing (0,66,2) with x^2, r2=0.0\n",
|
|
"fixing (0,67,0) with x^2, r2=0.9999999655544946\n",
|
|
"fixing (0,67,1) with x^2, r2=0.9999995390688572\n",
|
|
"fixing (0,67,2) with x^2, r2=0.9999997366108699\n",
|
|
"fixing (0,68,0) with x^2, r2=0.9999999735303753\n",
|
|
"fixing (0,68,1) with x^2, r2=0.9999999539372727\n",
|
|
"fixing (0,68,2) with x^2, r2=0.9999998409922631\n",
|
|
"fixing (0,69,0) with x^2, r2=0.9999999975190795\n",
|
|
"fixing (0,69,1) with x^2, r2=0.9999998840699803\n",
|
|
"fixing (0,69,2) with x^2, r2=0.9999999748333692\n",
|
|
"fixing (0,70,0) with x^2, r2=0.9999999638112955\n",
|
|
"fixing (0,70,1) with x^2, r2=0.999999996122007\n",
|
|
"fixing (0,70,2) with x^2, r2=0.9999990113519382\n",
|
|
"fixing (0,71,0) with x^2, r2=0.0\n",
|
|
"fixing (0,71,1) with x^2, r2=0.0\n",
|
|
"fixing (0,71,2) with x^2, r2=0.0\n",
|
|
"fixing (0,72,0) with x^2, r2=0.9999999782223539\n",
|
|
"fixing (0,72,1) with x^2, r2=0.9999996360566132\n",
|
|
"fixing (0,72,2) with x^2, r2=0.9999994783563169\n",
|
|
"fixing (0,73,0) with x^2, r2=0.0\n",
|
|
"fixing (0,73,1) with x^2, r2=0.0\n",
|
|
"fixing (0,73,2) with x^2, r2=0.0\n",
|
|
"fixing (0,74,0) with x^2, r2=0.9999999430582801\n",
|
|
"fixing (0,74,1) with x^2, r2=0.9999999373180665\n",
|
|
"fixing (0,74,2) with x^2, r2=0.9999999928808172\n",
|
|
"fixing (0,75,0) with x^2, r2=0.9999999675795376\n",
|
|
"fixing (0,75,1) with x^2, r2=0.9999999926331626\n",
|
|
"fixing (0,75,2) with x^2, r2=0.9999999455360133\n",
|
|
"fixing (0,76,0) with x^2, r2=0.9999999894203153\n",
|
|
"fixing (0,76,1) with x^2, r2=0.999999852706142\n",
|
|
"fixing (0,76,2) with x^2, r2=0.9999994569257162\n",
|
|
"fixing (0,77,0) with x^2, r2=0.0\n",
|
|
"fixing (0,77,1) with x^2, r2=0.0\n",
|
|
"fixing (0,77,2) with x^2, r2=0.0\n",
|
|
"fixing (0,78,0) with x^2, r2=0.9999969548814738\n",
|
|
"fixing (0,78,1) with x^2, r2=0.999999895396509\n",
|
|
"fixing (0,78,2) with x^2, r2=0.9999997624575255\n",
|
|
"fixing (0,79,0) with x^2, r2=0.0\n",
|
|
"fixing (0,79,1) with x^2, r2=0.0\n",
|
|
"fixing (0,79,2) with x^2, r2=0.0\n",
|
|
"fixing (0,80,0) with x^2, r2=0.0\n",
|
|
"fixing (0,80,1) with x^2, r2=0.0\n",
|
|
"fixing (0,80,2) with x^2, r2=0.0\n",
|
|
"fixing (0,81,0) with x^2, r2=0.9999999633167932\n",
|
|
"fixing (0,81,1) with x^2, r2=0.9999999924423665\n",
|
|
"fixing (0,81,2) with x^2, r2=0.9999999407891473\n",
|
|
"fixing (0,82,0) with x^2, r2=0.0\n",
|
|
"fixing (0,82,1) with x^2, r2=0.0\n",
|
|
"fixing (0,82,2) with x^2, r2=0.0\n",
|
|
"fixing (0,83,0) with x^2, r2=0.9964873061598577\n",
|
|
"fixing (0,83,1) with x^2, r2=0.9999998536697641\n",
|
|
"fixing (0,83,2) with x^2, r2=0.9999999474125241\n",
|
|
"fixing (0,84,0) with x^2, r2=0.9999999434524759\n",
|
|
"fixing (0,84,1) with x^2, r2=0.9999999848500863\n",
|
|
"fixing (0,84,2) with x^2, r2=0.9999997362933968\n",
|
|
"fixing (0,85,0) with x^2, r2=0.9999784391692933\n",
|
|
"fixing (0,85,1) with x^2, r2=0.9999999123872062\n",
|
|
"fixing (0,85,2) with x^2, r2=0.9999981066188347\n",
|
|
"fixing (0,86,0) with x^2, r2=0.9999999470214042\n",
|
|
"fixing (0,86,1) with x^2, r2=0.9999999622653485\n",
|
|
"fixing (0,86,2) with x^2, r2=0.9999999256587131\n",
|
|
"fixing (0,87,0) with x^2, r2=0.9999838246792585\n",
|
|
"fixing (0,87,1) with x^2, r2=0.9999998906573028\n",
|
|
"fixing (0,87,2) with x^2, r2=0.9997398325048757\n",
|
|
"fixing (0,88,0) with x^2, r2=0.9999903305520499\n",
|
|
"fixing (0,88,1) with x^2, r2=0.9999999129937596\n",
|
|
"fixing (0,88,2) with x^2, r2=0.9999994338574667\n",
|
|
"fixing (0,89,0) with x^2, r2=0.9999999969824458\n",
|
|
"fixing (0,89,1) with x^2, r2=0.9999998811902262\n",
|
|
"fixing (0,89,2) with x^2, r2=0.9999999955608072\n",
|
|
"fixing (0,90,0) with x^2, r2=0.9999999968821633\n",
|
|
"fixing (0,90,1) with x^2, r2=0.9999999231999729\n",
|
|
"fixing (0,90,2) with x^2, r2=0.999999921201756\n",
|
|
"fixing (0,91,0) with x^2, r2=0.9999734544061402\n",
|
|
"fixing (0,91,1) with x^2, r2=0.9999966985161072\n",
|
|
"fixing (0,91,2) with x^2, r2=0.9999999489971586\n",
|
|
"fixing (0,92,0) with x^2, r2=0.9999999864791468\n",
|
|
"fixing (0,92,1) with x^2, r2=0.9999999698743414\n",
|
|
"fixing (0,92,2) with x^2, r2=0.9998985820640515\n",
|
|
"fixing (0,93,0) with x^2, r2=0.0\n",
|
|
"fixing (0,93,1) with x^2, r2=0.0\n",
|
|
"fixing (0,93,2) with x^2, r2=0.0\n",
|
|
"fixing (0,94,0) with x^2, r2=0.9999572021042229\n",
|
|
"fixing (0,94,1) with x^2, r2=0.9999999403042822\n",
|
|
"fixing (0,94,2) with x^2, r2=0.9999984955483119\n",
|
|
"fixing (0,95,0) with x^2, r2=0.0\n",
|
|
"fixing (0,95,1) with x^2, r2=0.0\n",
|
|
"fixing (0,95,2) with x^2, r2=0.0\n",
|
|
"fixing (0,96,0) with x^2, r2=0.0\n",
|
|
"fixing (0,96,1) with x^2, r2=0.0\n",
|
|
"fixing (0,96,2) with x^2, r2=0.0\n",
|
|
"fixing (0,97,0) with x^2, r2=0.9999999855742208\n",
|
|
"fixing (0,97,1) with x^2, r2=0.9999990622913814\n",
|
|
"fixing (0,97,2) with x^2, r2=0.9999999661558678\n",
|
|
"fixing (0,98,0) with x^2, r2=0.9999998924577429\n",
|
|
"fixing (0,98,1) with x^2, r2=0.9999999075025128\n",
|
|
"fixing (0,98,2) with x^2, r2=0.9999925555905432\n",
|
|
"fixing (0,99,0) with x^2, r2=0.0\n",
|
|
"fixing (0,99,1) with x^2, r2=0.0\n",
|
|
"fixing (0,99,2) with x^2, r2=0.0\n",
|
|
"fixing (0,100,0) with x^2, r2=0.9999999888884751\n",
|
|
"fixing (0,100,1) with x^2, r2=0.9999999053398424\n",
|
|
"fixing (0,100,2) with x^2, r2=0.9999999274642732\n",
|
|
"fixing (0,101,0) with x^2, r2=0.0\n",
|
|
"fixing (0,101,1) with x^2, r2=0.0\n",
|
|
"fixing (0,101,2) with x^2, r2=0.0\n",
|
|
"fixing (0,102,0) with x^2, r2=0.0\n",
|
|
"fixing (0,102,1) with x^2, r2=0.0\n",
|
|
"fixing (0,102,2) with x^2, r2=0.0\n",
|
|
"fixing (0,103,0) with x^2, r2=0.9999997998513549\n",
|
|
"fixing (0,103,1) with x^2, r2=0.9999999874737161\n",
|
|
"fixing (0,103,2) with x^2, r2=0.9999999891891058\n",
|
|
"fixing (0,104,0) with x^2, r2=0.0\n",
|
|
"fixing (0,104,1) with x^2, r2=0.0\n",
|
|
"fixing (0,104,2) with x^2, r2=0.0\n",
|
|
"fixing (1,0,0) with x^2, r2=0.9827286380576173\n",
|
|
"fixing (1,0,1) with x^2, r2=0.9753307156038028\n",
|
|
"fixing (1,1,0) with x^2, r2=0.99206369703365\n",
|
|
"fixing (1,1,1) with x^2, r2=0.9950033104451041\n",
|
|
"fixing (1,2,0) with x^2, r2=0.9980758555730187\n",
|
|
"fixing (1,2,1) with x^2, r2=0.9973139539011773\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"lib = ['x','x^2']\n",
|
|
"\n",
|
|
"model.auto_symbolic(lib=lib)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 12,
|
|
"id": "a62fc07c-1522-4425-8f99-9ab673943cf1",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/latex": [
|
|
"$\\displaystyle 0.44 \\left(0.02 \\left(- x_{1} - 1\\right)^{2} + 0.02 \\left(x_{10} + 1\\right)^{2} + 0.04 \\left(- x_{101} - 1\\right)^{2} + 0.01 \\left(- x_{13} - 1\\right)^{2} - 0.02 \\left(- x_{14} - 1\\right)^{2} - 0.02 \\left(- x_{15} - 1\\right)^{2} + 0.02 \\left(- x_{17} - 1\\right)^{2} + 0.03 \\left(x_{2} + 1\\right)^{2} - 0.01 \\left(x_{20} + 1\\right)^{2} - 0.01 \\left(x_{21} + 1\\right)^{2} - 0.03 \\left(- x_{24} - 1\\right)^{2} + 0.01 \\left(- x_{26} - 1\\right)^{2} - 0.02 \\left(- x_{29} - 1\\right)^{2} - 0.02 \\left(- x_{31} - 1\\right)^{2} + 0.01 \\left(x_{32} + 1\\right)^{2} + 0.01 \\left(- x_{33} - 1\\right)^{2} - 0.01 \\left(x_{37} + 1\\right)^{2} - 0.01 \\left(- x_{39} - 1\\right)^{2} - 0.01 \\left(- x_{40} - 1\\right)^{2} - 0.02 \\left(- x_{54} - 1\\right)^{2} + 0.02 \\left(- x_{58} - 1\\right)^{2} - 0.01 \\left(- x_{6} - 1\\right)^{2} - 0.01 \\left(- x_{66} - 1\\right)^{2} - 0.02 \\left(- x_{68} - 1\\right)^{2} + 0.02 \\left(- x_{69} - 1\\right)^{2} - 0.04 \\left(x_{70} + 1\\right)^{2} + 0.01 \\left(- x_{71} - 1\\right)^{2} + 0.03 \\left(- x_{73} - 1\\right)^{2} + 0.01 \\left(- x_{75} - 1\\right)^{2} + 0.01 \\left(- x_{76} - 1\\right)^{2} + 0.02 \\left(- x_{77} - 1\\right)^{2} - 0.01 \\left(- x_{82} - 1\\right)^{2} - 0.01 \\left(- x_{85} - 1\\right)^{2} - 0.02 \\left(x_{87} + 1\\right)^{2} - 0.01 \\left(x_{9} + 1\\right)^{2} - 0.04 \\left(x_{90} + 1\\right)^{2} + 0.03 \\left(- x_{91} - 1\\right)^{2} + 0.02 \\left(x_{93} + 1\\right)^{2} + 0.03 \\left(x_{98} + 1\\right)^{2} - 0.01 \\left(- x_{99} - 1\\right)^{2} - 1\\right)^{2} + 0.7 \\left(- 0.03 \\left(- x_{1} - 1\\right)^{2} - 0.02 \\left(x_{10} + 1\\right)^{2} + 0.02 \\left(x_{101} + 1\\right)^{2} - 0.03 \\left(x_{104} + 1\\right)^{2} + 0.05 \\left(- x_{13} - 1\\right)^{2} + 0.01 \\left(- x_{15} - 1\\right)^{2} - 0.05 \\left(x_{16} + 1\\right)^{2} - 0.02 \\left(- x_{17} - 1\\right)^{2} - 0.01 \\left(- x_{2} - 1\\right)^{2} + 0.01 \\left(- x_{21} - 1\\right)^{2} + 0.02 \\left(x_{24} + 1\\right)^{2} - 0.01 \\left(- x_{26} - 1\\right)^{2} + 0.01 \\left(- x_{27} - 1\\right)^{2} - 0.02 \\left(- x_{28} - 1\\right)^{2} - 0.03 \\left(- x_{29} - 1\\right)^{2} + 0.03 \\left(- x_{3} - 1\\right)^{2} + 0.04 \\left(- x_{31} - 1\\right)^{2} + 0.05 \\left(- x_{32} - 1\\right)^{2} + 0.03 \\left(- x_{34} - 1\\right)^{2} - 0.01 \\left(- x_{37} - 1\\right)^{2} + 0.02 \\left(- x_{39} - 1\\right)^{2} - 0.03 \\left(x_{40} + 1\\right)^{2} - 0.02 \\left(x_{41} + 1\\right)^{2} - 0.07 \\left(- x_{54} - 1\\right)^{2} + 0.09 \\left(- x_{58} - 1\\right)^{2} + 0.03 \\left(x_{6} + 1\\right)^{2} - 0.02 \\left(- x_{66} - 1\\right)^{2} - 0.01 \\left(x_{68} + 1\\right)^{2} + 0.02 \\left(- x_{69} - 1\\right)^{2} - 0.03 \\left(x_{7} + 1\\right)^{2} + 0.02 \\left(x_{70} + 1\\right)^{2} - 0.01 \\left(x_{73} + 1\\right)^{2} + 0.04 \\left(x_{75} + 1\\right)^{2} + 0.01 \\left(x_{76} + 1\\right)^{2} - 0.01 \\left(x_{79} + 1\\right)^{2} + 0.01 \\left(- x_{82} - 1\\right)^{2} + 0.03 \\left(- x_{84} - 1\\right)^{2} + 0.01 \\left(x_{85} + 1\\right)^{2} + 0.02 \\left(- x_{87} - 1\\right)^{2} + 0.01 \\left(x_{89} + 1\\right)^{2} + 0.05 \\left(- x_{90} - 1\\right)^{2} - 0.01 \\left(- x_{91} - 1\\right)^{2} - 0.03 \\left(x_{92} + 1\\right)^{2} + 0.01 \\left(- x_{95} - 1\\right)^{2} + 0.03 \\left(- x_{98} - 1\\right)^{2} - 1\\right)^{2} + 0.17 \\left(- 0.01 \\left(- x_{1} - 1\\right)^{2} + 0.05 \\left(- x_{101} - 1\\right)^{2} - 0.07 \\left(x_{104} + 1\\right)^{2} + 0.06 \\left(- x_{14} - 1\\right)^{2} + 0.01 \\left(- x_{15} - 1\\right)^{2} + 0.02 \\left(- x_{16} - 1\\right)^{2} + 0.02 \\left(- x_{17} - 1\\right)^{2} + 0.02 \\left(- x_{20} - 1\\right)^{2} - 0.07 \\left(- x_{21} - 1\\right)^{2} + 0.05 \\left(x_{24} + 1\\right)^{2} + 0.05 \\left(- x_{26} - 1\\right)^{2} - 0.06 \\left(- x_{27} - 1\\right)^{2} - 0.01 \\left(- x_{28} - 1\\right)^{2} - 0.02 \\left(- x_{29} - 1\\right)^{2} - 0.02 \\left(x_{3} + 1\\right)^{2} + 0.06 \\left(- x_{31} - 1\\right)^{2} + 0.01 \\left(- x_{32} - 1\\right)^{2} + 0.05 \\left(- x_{34} - 1\\right)^{2} + 0.06 \\left(- x_{37} - 1\\right)^{2} + 0.03 \\left(- x_{38} - 1\\right)^{2} + 0.01 \\left(- x_{39} - 1\\right)^{2} - 0.13 \\left(- x_{54} - 1\\right)^{2} + 0.09 \\left(- x_{58} - 1\\right)^{2} - 0.04 \\left(x_{6} + 1\\right)^{2} + 0.02 \\left(x_{68} + 1\\right)^{2} + 0.07 \\left(x_{69} + 1\\right)^{2} + 0.04 \\left(- x_{7} - 1\\right)^{2} - 0.02 \\left(- x_{70} - 1\\right)^{2} + 0.08 \\left(- x_{71} - 1\\right)^{2} + 0.02 \\left(- x_{73} - 1\\right)^{2} + 0.03 \\left(- x_{75} - 1\\right)^{2} - 0.06 \\left(- x_{76} - 1\\right)^{2} + 0.02 \\left(- x_{77} - 1\\right)^{2} - 0.04 \\left(x_{79} + 1\\right)^{2} - 0.08 \\left(x_{82} + 1\\right)^{2} - 0.04 \\left(x_{84} + 1\\right)^{2} + 0.06 \\left(x_{85} + 1\\right)^{2} + 0.05 \\left(- x_{86} - 1\\right)^{2} + 0.07 \\left(- x_{87} - 1\\right)^{2} + 0.04 \\left(x_{88} + 1\\right)^{2} - 0.05 \\left(- x_{89} - 1\\right)^{2} + 0.12 \\left(x_{9} + 1\\right)^{2} - 0.02 \\left(x_{90} + 1\\right)^{2} - 0.02 \\left(- x_{91} - 1\\right)^{2} - 0.01 \\left(- x_{92} - 1\\right)^{2} - 0.04 \\left(- x_{93} - 1\\right)^{2} - 0.06 \\left(- x_{95} - 1\\right)^{2} + 0.01 \\left(x_{98} + 1\\right)^{2} - 0.05 \\left(- x_{99} - 1\\right)^{2} - 1\\right)^{2} - 0.57$"
|
|
],
|
|
"text/plain": [
|
|
"0.44*(0.02*(-x_1 - 1)**2 + 0.02*(x_10 + 1)**2 + 0.04*(-x_101 - 1)**2 + 0.01*(-x_13 - 1)**2 - 0.02*(-x_14 - 1)**2 - 0.02*(-x_15 - 1)**2 + 0.02*(-x_17 - 1)**2 + 0.03*(x_2 + 1)**2 - 0.e-2*(x_20 + 1)**2 - 0.e-2*(x_21 + 1)**2 - 0.03*(-x_24 - 1)**2 + 0.01*(-x_26 - 1)**2 - 0.02*(-x_29 - 1)**2 - 0.02*(-x_31 - 1)**2 + 0.01*(x_32 + 1)**2 + 0.01*(-x_33 - 1)**2 - 0.e-2*(x_37 + 1)**2 - 0.01*(-x_39 - 1)**2 - 0.e-2*(-x_40 - 1)**2 - 0.02*(-x_54 - 1)**2 + 0.02*(-x_58 - 1)**2 - 0.01*(-x_6 - 1)**2 - 0.01*(-x_66 - 1)**2 - 0.02*(-x_68 - 1)**2 + 0.02*(-x_69 - 1)**2 - 0.04*(x_70 + 1)**2 + 0.01*(-x_71 - 1)**2 + 0.03*(-x_73 - 1)**2 + 0.01*(-x_75 - 1)**2 + 0.01*(-x_76 - 1)**2 + 0.02*(-x_77 - 1)**2 - 0.01*(-x_82 - 1)**2 - 0.e-2*(-x_85 - 1)**2 - 0.02*(x_87 + 1)**2 - 0.e-2*(x_9 + 1)**2 - 0.04*(x_90 + 1)**2 + 0.03*(-x_91 - 1)**2 + 0.02*(x_93 + 1)**2 + 0.03*(x_98 + 1)**2 - 0.01*(-x_99 - 1)**2 - 1)**2 + 0.7*(-0.03*(-x_1 - 1)**2 - 0.02*(x_10 + 1)**2 + 0.02*(x_101 + 1)**2 - 0.03*(x_104 + 1)**2 + 0.05*(-x_13 - 1)**2 + 0.01*(-x_15 - 1)**2 - 0.05*(x_16 + 1)**2 - 0.02*(-x_17 - 1)**2 - 0.e-2*(-x_2 - 1)**2 + 0.01*(-x_21 - 1)**2 + 0.02*(x_24 + 1)**2 - 0.01*(-x_26 - 1)**2 + 0.01*(-x_27 - 1)**2 - 0.02*(-x_28 - 1)**2 - 0.03*(-x_29 - 1)**2 + 0.03*(-x_3 - 1)**2 + 0.04*(-x_31 - 1)**2 + 0.05*(-x_32 - 1)**2 + 0.03*(-x_34 - 1)**2 - 0.01*(-x_37 - 1)**2 + 0.02*(-x_39 - 1)**2 - 0.03*(x_40 + 1)**2 - 0.02*(x_41 + 1)**2 - 0.07*(-x_54 - 1)**2 + 0.09*(-x_58 - 1)**2 + 0.03*(x_6 + 1)**2 - 0.02*(-x_66 - 1)**2 - 0.01*(x_68 + 1)**2 + 0.02*(-x_69 - 1)**2 - 0.03*(x_7 + 1)**2 + 0.02*(x_70 + 1)**2 - 0.e-2*(x_73 + 1)**2 + 0.04*(x_75 + 1)**2 + 0.01*(x_76 + 1)**2 - 0.01*(x_79 + 1)**2 + 0.01*(-x_82 - 1)**2 + 0.03*(-x_84 - 1)**2 + 0.01*(x_85 + 1)**2 + 0.02*(-x_87 - 1)**2 + 0.01*(x_89 + 1)**2 + 0.05*(-x_90 - 1)**2 - 0.e-2*(-x_91 - 1)**2 - 0.03*(x_92 + 1)**2 + 0.01*(-x_95 - 1)**2 + 0.03*(-x_98 - 1)**2 - 1)**2 + 0.17*(-0.e-2*(-x_1 - 1)**2 + 0.05*(-x_101 - 1)**2 - 0.07*(x_104 + 1)**2 + 0.06*(-x_14 - 1)**2 + 0.01*(-x_15 - 1)**2 + 0.02*(-x_16 - 1)**2 + 0.02*(-x_17 - 1)**2 + 0.02*(-x_20 - 1)**2 - 0.07*(-x_21 - 1)**2 + 0.05*(x_24 + 1)**2 + 0.05*(-x_26 - 1)**2 - 0.06*(-x_27 - 1)**2 - 0.01*(-x_28 - 1)**2 - 0.02*(-x_29 - 1)**2 - 0.02*(x_3 + 1)**2 + 0.06*(-x_31 - 1)**2 + 0.01*(-x_32 - 1)**2 + 0.05*(-x_34 - 1)**2 + 0.06*(-x_37 - 1)**2 + 0.03*(-x_38 - 1)**2 + 0.01*(-x_39 - 1)**2 - 0.13*(-x_54 - 1)**2 + 0.09*(-x_58 - 1)**2 - 0.04*(x_6 + 1)**2 + 0.02*(x_68 + 1)**2 + 0.07*(x_69 + 1)**2 + 0.04*(-x_7 - 1)**2 - 0.02*(-x_70 - 1)**2 + 0.08*(-x_71 - 1)**2 + 0.02*(-x_73 - 1)**2 + 0.03*(-x_75 - 1)**2 - 0.06*(-x_76 - 1)**2 + 0.02*(-x_77 - 1)**2 - 0.04*(x_79 + 1)**2 - 0.08*(x_82 + 1)**2 - 0.04*(x_84 + 1)**2 + 0.06*(x_85 + 1)**2 + 0.05*(-x_86 - 1)**2 + 0.07*(-x_87 - 1)**2 + 0.04*(x_88 + 1)**2 - 0.05*(-x_89 - 1)**2 + 0.12*(x_9 + 1)**2 - 0.02*(x_90 + 1)**2 - 0.02*(-x_91 - 1)**2 - 0.e-2*(-x_92 - 1)**2 - 0.04*(-x_93 - 1)**2 - 0.06*(-x_95 - 1)**2 + 0.01*(x_98 + 1)**2 - 0.05*(-x_99 - 1)**2 - 1)**2 - 0.57"
|
|
]
|
|
},
|
|
"execution_count": 12,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"formula1, formula2 = model.symbolic_formula()[0]\n",
|
|
"formula1"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 13,
|
|
"id": "5051c2c0-772f-4b6f-b4f0-aa15cf7b6fe0",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"image/png": "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",
|
|
"text/plain": [
|
|
"<Figure size 500x400 with 322 Axes>"
|
|
]
|
|
},
|
|
"metadata": {},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"model.plot()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "1f86e6c9-1896-478f-ac95-c7b73ae6c28d",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": []
|
|
}
|
|
],
|
|
"metadata": {
|
|
"kernelspec": {
|
|
"display_name": "Python 3 (ipykernel)",
|
|
"language": "python",
|
|
"name": "python3"
|
|
},
|
|
"language_info": {
|
|
"codemirror_mode": {
|
|
"name": "ipython",
|
|
"version": 3
|
|
},
|
|
"file_extension": ".py",
|
|
"mimetype": "text/x-python",
|
|
"name": "python",
|
|
"nbconvert_exporter": "python",
|
|
"pygments_lexer": "ipython3",
|
|
"version": "3.9.7"
|
|
}
|
|
},
|
|
"nbformat": 4,
|
|
"nbformat_minor": 5
|
|
}
|