{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd\n",
"import matplotlib.pyplot as plt\n",
"from pathlib import Path\n",
"import numpy as np"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"path = 'C:/Users/diane/Desktop/xcalxcal.csv'\n",
"df = pd.read_csv(path,decimal = '.', sep = \";\", index_col=0)"
]
},
{
"cell_type": "code",
"execution_count": 32,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" name\n",
"cluster#2 s126\n",
"cluster#2 s256\n",
"cluster#1 s27\n",
"cluster#2 s166\n",
"cluster#2 s27\n",
"... ...\n",
"cluster#2 s356\n",
"cluster#2 s357\n",
"cluster#1 s358\n",
"cluster#2 s359\n",
"cluster#1 s360\n",
"\n",
"[361 rows x 1 columns]\n"
]
},
{
"ename": "ValueError",
"evalue": "not enough values to unpack (expected 2, got 1)",
"output_type": "error",
"traceback": [
"\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[1;31mValueError\u001b[0m Traceback (most recent call last)",
"Cell \u001b[1;32mIn[32], line 87\u001b[0m\n\u001b[0;32m 84\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m result\n\u001b[0;32m 86\u001b[0m m\u001b[38;5;241m=\u001b[39m SkKmeans()\n\u001b[1;32m---> 87\u001b[0m a,b \u001b[38;5;241m=\u001b[39m m\u001b[38;5;241m.\u001b[39mkmeans1layer(x \u001b[38;5;241m=\u001b[39m df, ratio \u001b[38;5;241m=\u001b[39m \u001b[38;5;241m0.2\u001b[39m, max_k\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m10\u001b[39m)\n",
"\u001b[1;31mValueError\u001b[0m: not enough values to unpack (expected 2, got 1)"
]
}
],
"source": [
" def kmeans(x, max_k):\n",
" k = Cluster.find_optimal_k(X=x)\n",
" model = KMeans(n_clusters=k, init='k-means++', max_iter=300, n_init=10, random_state=42)\n",
" model.fit(x)\n",
" clu = [f'cluster#{i}' for i in model.predict(x)+1]\n",
" res = tuple(zip(clu, x.index))\n",
" centers = model.cluster_centers_\n",
" for i in set(clu):\n",
" # search the closest points of the cluster members to center of the cluster\n",
" medoids[i], _ = pairwise_distances_argmin_min(tcr.iloc[clustered,:], clu_centers)\n",
"\n",
" return res, medoids"
]
},
{
"cell_type": "code",
"execution_count": 288,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"{'cluster#1': 's315', 'cluster#2': 's303'}"
]
},
"execution_count": 288,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"\n",
" \n",
"\n",
"M = SkKmeans()\n",
"res, medoids = M.kmeans(df, max_k=40)\n",
"medoids"
]
},
{
"cell_type": "code",
"execution_count": 309,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>names</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>cluster#2</th>\n",
" <td>s126</td>\n",
" </tr>\n",
" <tr>\n",
" <th>cluster#2</th>\n",
" <td>s256</td>\n",
" </tr>\n",
" <tr>\n",
" <th>cluster#1</th>\n",
" <td>s27</td>\n",
" </tr>\n",
" <tr>\n",
" <th>cluster#2</th>\n",
" <td>s166</td>\n",
" </tr>\n",
" <tr>\n",
" <th>cluster#2</th>\n",
" <td>s27</td>\n",
" </tr>\n",
" <tr>\n",
" <th>...</th>\n",
" <td>...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>cluster#2</th>\n",
" <td>s356</td>\n",
" </tr>\n",
" <tr>\n",
" <th>cluster#2</th>\n",
" <td>s357</td>\n",
" </tr>\n",
" <tr>\n",
" <th>cluster#1</th>\n",
" <td>s358</td>\n",
" </tr>\n",
" <tr>\n",
" <th>cluster#2</th>\n",
" <td>s359</td>\n",
" </tr>\n",
" <tr>\n",
" <th>cluster#1</th>\n",
" <td>s360</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>361 rows × 1 columns</p>\n",
"</div>"
],
"text/plain": [
" names\n",
"cluster#2 s126\n",
"cluster#2 s256\n",
"cluster#1 s27\n",
"cluster#2 s166\n",
"cluster#2 s27\n",
"... ...\n",
"cluster#2 s356\n",
"cluster#2 s357\n",
"cluster#1 s358\n",
"cluster#2 s359\n",
"cluster#1 s360\n",
"\n",
"[361 rows x 1 columns]"
]
},
"execution_count": 309,
"metadata": {},
"output_type": "execute_result"
}
],
"source": []
},
{
"cell_type": "code",
"execution_count": 312,
"metadata": {},
"outputs": [],
"source": [
"X = df"
]
},
{
"cell_type": "code",
"execution_count": 430,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['s116', 's212']"
]
},
"execution_count": 430,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"def selection_method(X, method, **kwargs):\n",
" #['random', 'kennard-stone', 'medoids', 'meta-clusters']\n",
" if method =='random':\n",
" from sklearn.model_selection import train_test_split\n",
" elif method == 'kennard-stone':\n",
" from kennard_stone import train_test_split\n",
" if method in ['random','kennard-stone']:\n",
" selected, _ = train_test_split(X, train_size= kwargs['rset'])\n",
" sname = selected.index\n",
"\n",
" if method in ['meta-ks','meta-medoids']:\n",
" best_k = 2\n",
" best_score = -1\n",
" for k in range(2, min(10,X.shape[0])):\n",
" from sklearn.cluster import KMeans\n",
" from sklearn.metrics import silhouette_score\n",
" model = KMeans(n_clusters=best_k, random_state=42, init='random', n_init=1, max_iter=100)\n",
" labels = model.fit_predict(X)\n",
" score = silhouette_score(X, labels)\n",
" if score > best_score:\n",
" best_score = score\n",
" best_k = k \n",
" model = KMeans(n_clusters=best_k, random_state=42, init='random', n_init=1, max_iter=100)\n",
" model.fit(X)\n",
" yp = model.predict(X)\n",
"\n",
" sname = []\n",
" for i in range(best_k):\n",
" t = X.loc[yp==i]\n",
" if method == \"meta-medoids\":\n",
" from scipy.spatial.distance import cdist\n",
" distances = cdist(t.values, t.values, metric='euclidean') \n",
" sum_distances = np.sum(distances, axis=1)\n",
" medoid_index = np.argmin(sum_distances)\n",
" sname.append(X.index[medoid_index])\n",
" \n",
" elif method == 'meta-ks':\n",
" from kennard_stone import train_test_split\n",
" selected, _ = train_test_split(t, train_size= kwargs['rset_meta'])\n",
" sname.append(selected.index)\n",
" return sname\n",
"l = ['random', 'kennard-stone', 'meta-medoids', 'meta-ks']\n",
"selection_method(X=df, method= l[2], rset = 0.01, rset_meta = 0.2)"
]
},
{
"cell_type": "code",
"execution_count": 55,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"4\n",
"1\n"
]
},
{
"data": {
"text/plain": [
"{}"
]
},
"execution_count": 55,
"metadata": {},
"output_type": "execute_result"
}
],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"mask = df.index.duplicated(keep=False) # Keep all duplicates (True for replicated)\n",
"\n",
"# For the duplicated sample_ids, apply suffix (_1, _2, etc.)\n",
"df.index = df.index.where(~mask, \n",
" df.groupby(df.index).cumcount().add(1).astype(str).radd(df.index + '#'))\n",
"len(set(df.index))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"x.T.plot(y=x.index, kind='line', legend=False, )"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"267 "
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"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": "3.12.7"
}
},
"nbformat": 4,
"nbformat_minor": 2
}