{
"cells": [
{
"cell_type": "markdown",
"id": "b69b2ab7",
"metadata": {},
"source": [
"# FateZ Clustering \n",
"\n",
"This notebook demonstrate how to implement clustering method with FateZ's representing method"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "dd050393",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"This part is yet to be modified!\n",
"Done Import\n"
]
}
],
"source": [
"print('This part is yet to be modified!')\n",
"\n",
"import os\n",
"import torch\n",
"import numpy as np\n",
"from torch.utils.data import DataLoader\n",
"from pkg_resources import resource_filename\n",
"from sklearn import cluster\n",
"import fatez.test as test\n",
"import fatez.model as model\n",
"# import scanpy as sc\n",
"\n",
"print('Done Import')"
]
},
{
"cell_type": "markdown",
"id": "10210c8d",
"metadata": {},
"source": [
"### Initialize testing model first."
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "3948e182",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Testing Full Model.\n",
"\n",
"\tPre-Trainer Green.\n",
"\n",
"\tFine-Tuner Green.\n",
"\n",
"Edge Explain:\n",
" tensor([[0.1573, 0.1524, 0.1506, 0.1357, 0.1349, 0.1365, 0.1361, 0.1353, 0.0000,\n",
" 0.1383],\n",
" [0.1358, 0.1370, 0.1243, 0.1215, 0.1255, 0.1237, 0.1196, 0.1621, 0.0000,\n",
" 0.1627],\n",
" [0.1638, 0.1475, 0.1416, 0.1444, 0.1399, 0.1464, 0.1395, 0.1368, 0.0000,\n",
" 0.1393],\n",
" [0.1305, 0.1263, 0.1264, 0.1260, 0.1198, 0.1493, 0.1356, 0.1604, 0.0000,\n",
" 0.1371]]) \n",
"\n",
"Reg Explain:\n",
" tensor([0., 0., 0., 0.], dtype=torch.float64) \n",
"\n",
"Node Explain:\n",
" tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0.], dtype=torch.float64) \n",
"\n",
"\tExplainer Green.\n",
"\n"
]
}
],
"source": [
"faker = test.Faker()\n",
"testM, _ = faker.test_full_model()\n",
"# model.Save(faker.test_gat(), '../data/ignore/gat.model')\n",
"# model.Save(testM, '../data/ignore/trainer.model')"
]
},
{
"cell_type": "markdown",
"id": "255aa807",
"metadata": {},
"source": [
"### Get the fake dataset"
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "82ce2fb9",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮\n",
"│ in <cell line: 2>:2 │\n",
"│ │\n",
"│ 1 dataset = faker.make_data_loader().dataset │\n",
"│ ❱ 2 for x in DataLoader(dataset, batch_size = len(dataset)): │\n",
"│ 3 │ all_fea_mat = x[0] │\n",
"│ 4 │ all_adj_mat = x[1] │\n",
"│ 5 print(f'Labels:\\n{labels.tolist()}') │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:634 in __next__ │\n",
"│ │\n",
"│ 631 │ │ │ if self._sampler_iter is None: │\n",
"│ 632 │ │ │ │ # TODO(https://github.com/pytorch/pytorch/issues/76750) │\n",
"│ 633 │ │ │ │ self._reset() # type: ignore[call-arg] │\n",
"│ ❱ 634 │ │ │ data = self._next_data() │\n",
"│ 635 │ │ │ self._num_yielded += 1 │\n",
"│ 636 │ │ │ if self._dataset_kind == _DatasetKind.Iterable and \\ │\n",
"│ 637 │ │ │ │ │ self._IterableDataset_len_called is not None and \\ │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:678 in _next_data │\n",
"│ │\n",
"│ 675 │ │\n",
"│ 676 │ def _next_data(self): │\n",
"│ 677 │ │ index = self._next_index() # may raise StopIteration │\n",
"│ ❱ 678 │ │ data = self._dataset_fetcher.fetch(index) # may raise StopIteration │\n",
"│ 679 │ │ if self._pin_memory: │\n",
"│ 680 │ │ │ data = _utils.pin_memory.pin_memory(data, self._pin_memory_device) │\n",
"│ 681 │ │ return data │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/torch/utils/data/_utils/fetch.py:54 in fetch │\n",
"│ │\n",
"│ 51 │ │ │ │ data = [self.dataset[idx] for idx in possibly_batched_index] │\n",
"│ 52 │ │ else: │\n",
"│ 53 │ │ │ data = self.dataset[possibly_batched_index] │\n",
"│ ❱ 54 │ │ return self.collate_fn(data) │\n",
"│ 55 │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/torch/utils/data/_utils/collate.py:264 in │\n",
"│ default_collate │\n",
"│ │\n",
"│ 261 │ │ │ >>> default_collate_fn_map.update(CustoType, collate_customtype_fn) │\n",
"│ 262 │ │ │ >>> default_collate(batch) # Handle `CustomType` automatically │\n",
"│ 263 │ \"\"\" │\n",
"│ ❱ 264 │ return collate(batch, collate_fn_map=default_collate_fn_map) │\n",
"│ 265 │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/torch/utils/data/_utils/collate.py:150 in collate │\n",
"│ │\n",
"│ 147 │ │ │ │ # The sequence type may not support `__init__(iterable)` (e.g., `range`) │\n",
"│ 148 │ │ │ │ return [collate(samples, collate_fn_map=collate_fn_map) for samples in t │\n",
"│ 149 │ │\n",
"│ ❱ 150 │ raise TypeError(default_collate_err_msg_format.format(elem_type)) │\n",
"│ 151 │\n",
"│ 152 │\n",
"│ 153 def collate_tensor_fn(batch, *, collate_fn_map: Optional[Dict[Union[Type, Tuple[Type, .. │\n",
"╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\n",
"TypeError: default_collate: batch must contain tensors, numpy arrays, numbers, dicts or lists; found <class \n",
"'torch_geometric.data.data.Data'>\n",
"\n"
],
"text/plain": [
"\u001b[31m╭─\u001b[0m\u001b[31m──────────────────────────────\u001b[0m\u001b[31m \u001b[0m\u001b[1;31mTraceback \u001b[0m\u001b[1;2;31m(most recent call last)\u001b[0m\u001b[31m \u001b[0m\u001b[31m───────────────────────────────\u001b[0m\u001b[31m─╮\u001b[0m\n",
"\u001b[31m│\u001b[0m in \u001b[92m\u001b[0m:\u001b[94m2\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m1 \u001b[0mdataset = faker.make_data_loader().dataset \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m❱ \u001b[0m2 \u001b[94mfor\u001b[0m x \u001b[95min\u001b[0m DataLoader(dataset, batch_size = \u001b[96mlen\u001b[0m(dataset)): \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m3 \u001b[0m\u001b[2m│ \u001b[0mall_fea_mat = x[\u001b[94m0\u001b[0m] \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m4 \u001b[0m\u001b[2m│ \u001b[0mall_adj_mat = x[\u001b[94m1\u001b[0m] \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m5 \u001b[0m\u001b[96mprint\u001b[0m(\u001b[33mf\u001b[0m\u001b[33m'\u001b[0m\u001b[33mLabels:\u001b[0m\u001b[33m\\n\u001b[0m\u001b[33m{\u001b[0mlabels.tolist()\u001b[33m}\u001b[0m\u001b[33m'\u001b[0m) \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2;33m/usr/local/lib/python3.10/dist-packages/torch/utils/data/\u001b[0m\u001b[1;33mdataloader.py\u001b[0m:\u001b[94m634\u001b[0m in \u001b[92m__next__\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 631 \u001b[0m\u001b[2m│ │ │ \u001b[0m\u001b[94mif\u001b[0m \u001b[96mself\u001b[0m._sampler_iter \u001b[95mis\u001b[0m \u001b[94mNone\u001b[0m: \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 632 \u001b[0m\u001b[2m│ │ │ │ \u001b[0m\u001b[2m# TODO(https://github.com/pytorch/pytorch/issues/76750)\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 633 \u001b[0m\u001b[2m│ │ │ │ \u001b[0m\u001b[96mself\u001b[0m._reset() \u001b[2m# type: ignore[call-arg]\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m❱ \u001b[0m 634 \u001b[2m│ │ │ \u001b[0mdata = \u001b[96mself\u001b[0m._next_data() \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 635 \u001b[0m\u001b[2m│ │ │ \u001b[0m\u001b[96mself\u001b[0m._num_yielded += \u001b[94m1\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 636 \u001b[0m\u001b[2m│ │ │ \u001b[0m\u001b[94mif\u001b[0m \u001b[96mself\u001b[0m._dataset_kind == _DatasetKind.Iterable \u001b[95mand\u001b[0m \\ \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 637 \u001b[0m\u001b[2m│ │ │ │ │ \u001b[0m\u001b[96mself\u001b[0m._IterableDataset_len_called \u001b[95mis\u001b[0m \u001b[95mnot\u001b[0m \u001b[94mNone\u001b[0m \u001b[95mand\u001b[0m \\ \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2;33m/usr/local/lib/python3.10/dist-packages/torch/utils/data/\u001b[0m\u001b[1;33mdataloader.py\u001b[0m:\u001b[94m678\u001b[0m in \u001b[92m_next_data\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 675 \u001b[0m\u001b[2m│ \u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 676 \u001b[0m\u001b[2m│ \u001b[0m\u001b[94mdef\u001b[0m \u001b[92m_next_data\u001b[0m(\u001b[96mself\u001b[0m): \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 677 \u001b[0m\u001b[2m│ │ \u001b[0mindex = \u001b[96mself\u001b[0m._next_index() \u001b[2m# may raise StopIteration\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m❱ \u001b[0m 678 \u001b[2m│ │ \u001b[0mdata = \u001b[96mself\u001b[0m._dataset_fetcher.fetch(index) \u001b[2m# may raise StopIteration\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 679 \u001b[0m\u001b[2m│ │ \u001b[0m\u001b[94mif\u001b[0m \u001b[96mself\u001b[0m._pin_memory: \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 680 \u001b[0m\u001b[2m│ │ │ \u001b[0mdata = _utils.pin_memory.pin_memory(data, \u001b[96mself\u001b[0m._pin_memory_device) \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 681 \u001b[0m\u001b[2m│ │ \u001b[0m\u001b[94mreturn\u001b[0m data \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2;33m/usr/local/lib/python3.10/dist-packages/torch/utils/data/_utils/\u001b[0m\u001b[1;33mfetch.py\u001b[0m:\u001b[94m54\u001b[0m in \u001b[92mfetch\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m51 \u001b[0m\u001b[2m│ │ │ │ \u001b[0mdata = [\u001b[96mself\u001b[0m.dataset[idx] \u001b[94mfor\u001b[0m idx \u001b[95min\u001b[0m possibly_batched_index] \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m52 \u001b[0m\u001b[2m│ │ \u001b[0m\u001b[94melse\u001b[0m: \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m53 \u001b[0m\u001b[2m│ │ │ \u001b[0mdata = \u001b[96mself\u001b[0m.dataset[possibly_batched_index] \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m❱ \u001b[0m54 \u001b[2m│ │ \u001b[0m\u001b[94mreturn\u001b[0m \u001b[96mself\u001b[0m.collate_fn(data) \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m55 \u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2;33m/usr/local/lib/python3.10/dist-packages/torch/utils/data/_utils/\u001b[0m\u001b[1;33mcollate.py\u001b[0m:\u001b[94m264\u001b[0m in \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[92mdefault_collate\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m261 \u001b[0m\u001b[2;33m│ │ │ \u001b[0m\u001b[33m>>> default_collate_fn_map.update(CustoType, collate_customtype_fn)\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m262 \u001b[0m\u001b[2;33m│ │ │ \u001b[0m\u001b[33m>>> default_collate(batch) # Handle `CustomType` automatically\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m263 \u001b[0m\u001b[2;33m│ \u001b[0m\u001b[33m\"\"\"\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m❱ \u001b[0m264 \u001b[2m│ \u001b[0m\u001b[94mreturn\u001b[0m collate(batch, collate_fn_map=default_collate_fn_map) \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m265 \u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2;33m/usr/local/lib/python3.10/dist-packages/torch/utils/data/_utils/\u001b[0m\u001b[1;33mcollate.py\u001b[0m:\u001b[94m150\u001b[0m in \u001b[92mcollate\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m147 \u001b[0m\u001b[2m│ │ │ │ \u001b[0m\u001b[2m# The sequence type may not support `__init__(iterable)` (e.g., `range`)\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m148 \u001b[0m\u001b[2m│ │ │ │ \u001b[0m\u001b[94mreturn\u001b[0m [collate(samples, collate_fn_map=collate_fn_map) \u001b[94mfor\u001b[0m samples \u001b[95min\u001b[0m t \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m149 \u001b[0m\u001b[2m│ \u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m❱ \u001b[0m150 \u001b[2m│ \u001b[0m\u001b[94mraise\u001b[0m \u001b[96mTypeError\u001b[0m(default_collate_err_msg_format.format(elem_type)) \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m151 \u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m152 \u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m153 \u001b[0m\u001b[94mdef\u001b[0m \u001b[92mcollate_tensor_fn\u001b[0m(batch, *, collate_fn_map: Optional[Dict[Union[Type, Tuple[Type, .. \u001b[31m│\u001b[0m\n",
"\u001b[31m╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\u001b[0m\n",
"\u001b[1;91mTypeError: \u001b[0mdefault_collate: batch must contain tensors, numpy arrays, numbers, dicts or lists; found \u001b[1m<\u001b[0m\u001b[1;95mclass\u001b[0m\u001b[39m \u001b[0m\n",
"\u001b[32m'torch_geometric.data.data.Data'\u001b[0m\u001b[1m>\u001b[0m\n"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"dataset = faker.make_data_loader().dataset\n",
"for x in DataLoader(dataset, batch_size = len(dataset)):\n",
" all_fea_mat = x[0]\n",
" all_adj_mat = x[1]\n",
"print(f'Labels:\\n{labels.tolist()}')"
]
},
{
"cell_type": "markdown",
"id": "f82982f2",
"metadata": {},
"source": [
"### Process origin data"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "59d0aef5",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮\n",
"│ in <cell line: 2>:2 │\n",
"│ │\n",
"│ 1 # Flatten Data │\n",
"│ ❱ 2 origin = np.array([torch.reshape(ele.to_dense(), (-1,)).tolist() for ele in all_fea_mat] │\n",
"│ 3 │\n",
"│ 4 # PCA analysis for dimensionality deduction │\n",
"│ 5 pca_analysis = sc.pp.pca(origin, n_comps = 9, return_info = True,) │\n",
"╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\n",
"NameError: name 'all_fea_mat' is not defined\n",
"\n"
],
"text/plain": [
"\u001b[31m╭─\u001b[0m\u001b[31m──────────────────────────────\u001b[0m\u001b[31m \u001b[0m\u001b[1;31mTraceback \u001b[0m\u001b[1;2;31m(most recent call last)\u001b[0m\u001b[31m \u001b[0m\u001b[31m───────────────────────────────\u001b[0m\u001b[31m─╮\u001b[0m\n",
"\u001b[31m│\u001b[0m in \u001b[92m\u001b[0m:\u001b[94m2\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m1 \u001b[0m\u001b[2m# Flatten Data\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m❱ \u001b[0m2 origin = np.array([torch.reshape(ele.to_dense(), (-\u001b[94m1\u001b[0m,)).tolist() \u001b[94mfor\u001b[0m ele \u001b[95min\u001b[0m all_fea_mat] \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m3 \u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m4 \u001b[0m\u001b[2m# PCA analysis for dimensionality deduction\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m5 \u001b[0mpca_analysis = sc.pp.pca(origin, n_comps = \u001b[94m9\u001b[0m, return_info = \u001b[94mTrue\u001b[0m,) \u001b[31m│\u001b[0m\n",
"\u001b[31m╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\u001b[0m\n",
"\u001b[1;91mNameError: \u001b[0mname \u001b[32m'all_fea_mat'\u001b[0m is not defined\n"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Flatten Data\n",
"origin = np.array([torch.reshape(ele.to_dense(), (-1,)).tolist() for ele in all_fea_mat])\n",
"\n",
"# PCA analysis for dimensionality deduction\n",
"pca_analysis = sc.pp.pca(origin, n_comps = 9, return_info = True,)\n",
"origin_pca = pca_analysis[0]\n",
"var_ratios = pca_analysis[2]\n",
"print(f'Origin Data Var Ratios:\\n{var_ratios}')\n"
]
},
{
"cell_type": "markdown",
"id": "96be974b",
"metadata": {},
"source": [
"### Process data with encoder"
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "d90f7dc8",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮\n",
"│ in <cell line: 2>:3 │\n",
"│ │\n",
"│ 1 # Get encoded representaions made by GAT -> BERT encoder │\n",
"│ 2 encode = np.array([ │\n",
"│ ❱ 3 │ torch.reshape(ele, (-1,)).tolist() for ele in testM.get_encoder_output( │\n",
"│ 4 │ │ all_fea_mat, all_adj_mat │\n",
"│ 5 │ ) │\n",
"│ 6 ]) │\n",
"╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\n",
"AttributeError: 'Trainer' object has no attribute 'get_encoder_output'\n",
"\n"
],
"text/plain": [
"\u001b[31m╭─\u001b[0m\u001b[31m──────────────────────────────\u001b[0m\u001b[31m \u001b[0m\u001b[1;31mTraceback \u001b[0m\u001b[1;2;31m(most recent call last)\u001b[0m\u001b[31m \u001b[0m\u001b[31m───────────────────────────────\u001b[0m\u001b[31m─╮\u001b[0m\n",
"\u001b[31m│\u001b[0m in \u001b[92m\u001b[0m:\u001b[94m3\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 1 \u001b[0m\u001b[2m# Get encoded representaions made by GAT -> BERT encoder\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 2 \u001b[0mencode = np.array([ \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m❱ \u001b[0m 3 \u001b[2m│ \u001b[0mtorch.reshape(ele, (-\u001b[94m1\u001b[0m,)).tolist() \u001b[94mfor\u001b[0m ele \u001b[95min\u001b[0m testM.get_encoder_output( \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 4 \u001b[0m\u001b[2m│ │ \u001b[0mall_fea_mat, all_adj_mat \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 5 \u001b[0m\u001b[2m│ \u001b[0m) \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 6 \u001b[0m]) \u001b[31m│\u001b[0m\n",
"\u001b[31m╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\u001b[0m\n",
"\u001b[1;91mAttributeError: \u001b[0m\u001b[32m'Trainer'\u001b[0m object has no attribute \u001b[32m'get_encoder_output'\u001b[0m\n"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Get encoded representaions made by GAT -> BERT encoder\n",
"encode = np.array([\n",
" torch.reshape(ele, (-1,)).tolist() for ele in testM.get_encoder_output(\n",
" all_fea_mat, all_adj_mat\n",
" )\n",
"])\n",
"\n",
"# PCA analysis for dimensionality deduction\n",
"pca_analysis = sc.pp.pca(encode, n_comps = 9, return_info = True,)\n",
"encode_pca = pca_analysis[0]\n",
"var_ratios = pca_analysis[2]\n",
"print(f'Encoded Rep Var Ratios:\\n{var_ratios}')"
]
},
{
"cell_type": "markdown",
"id": "b5299e3c",
"metadata": {},
"source": [
"### Set clustering models and fit models with original data"
]
},
{
"cell_type": "code",
"execution_count": 11,
"id": "48d8fea6",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮\n",
"│ in <cell line: 2>:2 │\n",
"│ │\n",
"│ 1 eps = 0.5 │\n",
"│ ❱ 2 n_clusters = len(np.unique(labels)) │\n",
"│ 3 min_samples = 5 │\n",
"│ 4 │\n",
"│ 5 dbscan = cluster.DBSCAN(eps = eps) │\n",
"╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\n",
"NameError: name 'labels' is not defined\n",
"\n"
],
"text/plain": [
"\u001b[31m╭─\u001b[0m\u001b[31m──────────────────────────────\u001b[0m\u001b[31m \u001b[0m\u001b[1;31mTraceback \u001b[0m\u001b[1;2;31m(most recent call last)\u001b[0m\u001b[31m \u001b[0m\u001b[31m───────────────────────────────\u001b[0m\u001b[31m─╮\u001b[0m\n",
"\u001b[31m│\u001b[0m in \u001b[92m\u001b[0m:\u001b[94m2\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 1 \u001b[0meps = \u001b[94m0.5\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m❱ \u001b[0m 2 n_clusters = \u001b[96mlen\u001b[0m(np.unique(labels)) \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 3 \u001b[0mmin_samples = \u001b[94m5\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 4 \u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 5 \u001b[0mdbscan = cluster.DBSCAN(eps = eps) \u001b[31m│\u001b[0m\n",
"\u001b[31m╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\u001b[0m\n",
"\u001b[1;91mNameError: \u001b[0mname \u001b[32m'labels'\u001b[0m is not defined\n"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"eps = 0.5\n",
"n_clusters = len(np.unique(labels))\n",
"min_samples = 5\n",
"\n",
"dbscan = cluster.DBSCAN(eps = eps)\n",
"kmeans = cluster.KMeans(n_clusters = n_clusters)\n",
"optics = cluster.OPTICS(min_samples = min_samples)\n",
"\n",
"dbscan.fit(origin_pca)\n",
"kmeans.fit(origin_pca)\n",
"optics.fit(origin_pca)\n",
"\n",
"# Get labels\n",
"print(dbscan.labels_.astype(int))\n",
"print(kmeans.labels_.astype(int))\n",
"print(optics.labels_.astype(int))"
]
},
{
"cell_type": "markdown",
"id": "eba48c10",
"metadata": {},
"source": [
"### Reset models and fit with encoded representaions"
]
},
{
"cell_type": "code",
"execution_count": 12,
"id": "1f1098d5",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮\n",
"│ in <cell line: 2>:2 │\n",
"│ │\n",
"│ 1 dbscan = cluster.DBSCAN(eps = eps) │\n",
"│ ❱ 2 kmeans = cluster.KMeans(n_clusters = n_clusters) │\n",
"│ 3 optics = cluster.OPTICS(min_samples = min_samples) │\n",
"│ 4 dbscan.fit(encode_pca) │\n",
"│ 5 kmeans.fit(encode_pca) │\n",
"╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\n",
"NameError: name 'n_clusters' is not defined\n",
"\n"
],
"text/plain": [
"\u001b[31m╭─\u001b[0m\u001b[31m──────────────────────────────\u001b[0m\u001b[31m \u001b[0m\u001b[1;31mTraceback \u001b[0m\u001b[1;2;31m(most recent call last)\u001b[0m\u001b[31m \u001b[0m\u001b[31m───────────────────────────────\u001b[0m\u001b[31m─╮\u001b[0m\n",
"\u001b[31m│\u001b[0m in \u001b[92m\u001b[0m:\u001b[94m2\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 1 \u001b[0mdbscan = cluster.DBSCAN(eps = eps) \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[31m❱ \u001b[0m 2 kmeans = cluster.KMeans(n_clusters = n_clusters) \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 3 \u001b[0moptics = cluster.OPTICS(min_samples = min_samples) \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 4 \u001b[0mdbscan.fit(encode_pca) \u001b[31m│\u001b[0m\n",
"\u001b[31m│\u001b[0m \u001b[2m 5 \u001b[0mkmeans.fit(encode_pca) \u001b[31m│\u001b[0m\n",
"\u001b[31m╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\u001b[0m\n",
"\u001b[1;91mNameError: \u001b[0mname \u001b[32m'n_clusters'\u001b[0m is not defined\n"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"dbscan = cluster.DBSCAN(eps = eps)\n",
"kmeans = cluster.KMeans(n_clusters = n_clusters)\n",
"optics = cluster.OPTICS(min_samples = min_samples)\n",
"dbscan.fit(encode_pca)\n",
"kmeans.fit(encode_pca)\n",
"optics.fit(encode_pca)\n",
"\n",
"# Get labels\n",
"print(dbscan.labels_.astype(int))\n",
"print(kmeans.labels_.astype(int))\n",
"print(optics.labels_.astype(int))"
]
}
],
"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": "3.10.6"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
| | | | |