Spaces:
Runtime error
Runtime error
File size: 18,695 Bytes
e331e72 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 |
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Visualizing the knowledge graph with `yfiles-jupyter-graphs`\n",
"\n",
"This notebook is a partial copy of [local_search.ipynb](../../local_search.ipynb) that shows how to use `yfiles-jupyter-graphs` to add interactive graph visualizations of the parquet files and how to visualize the result context of `graphrag` queries (see at the end of this notebook)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Copyright (c) 2024 Microsoft Corporation.\n",
"# Licensed under the MIT License."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"\n",
"import pandas as pd\n",
"import tiktoken\n",
"\n",
"from graphrag.query.context_builder.entity_extraction import EntityVectorStoreKey\n",
"from graphrag.query.indexer_adapters import (\n",
" read_indexer_covariates,\n",
" read_indexer_entities,\n",
" read_indexer_relationships,\n",
" read_indexer_reports,\n",
" read_indexer_text_units,\n",
")\n",
"from graphrag.query.input.loaders.dfs import (\n",
" store_entity_semantic_embeddings,\n",
")\n",
"from graphrag.query.llm.oai.chat_openai import ChatOpenAI\n",
"from graphrag.query.llm.oai.embedding import OpenAIEmbedding\n",
"from graphrag.query.llm.oai.typing import OpenaiApiType\n",
"from graphrag.query.structured_search.local_search.mixed_context import (\n",
" LocalSearchMixedContext,\n",
")\n",
"from graphrag.query.structured_search.local_search.search import LocalSearch\n",
"from graphrag.vector_stores.lancedb import LanceDBVectorStore"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Local Search Example\n",
"\n",
"Local search method generates answers by combining relevant data from the AI-extracted knowledge-graph with text chunks of the raw documents. This method is suitable for questions that require an understanding of specific entities mentioned in the documents (e.g. What are the healing properties of chamomile?)."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load text units and graph data tables as context for local search\n",
"\n",
"- In this test we first load indexing outputs from parquet files to dataframes, then convert these dataframes into collections of data objects aligning with the knowledge model."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load tables to dataframes"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"INPUT_DIR = \"../../inputs/operation dulce\"\n",
"LANCEDB_URI = f\"{INPUT_DIR}/lancedb\"\n",
"\n",
"COMMUNITY_REPORT_TABLE = \"create_final_community_reports\"\n",
"ENTITY_TABLE = \"create_final_nodes\"\n",
"ENTITY_EMBEDDING_TABLE = \"create_final_entities\"\n",
"RELATIONSHIP_TABLE = \"create_final_relationships\"\n",
"COVARIATE_TABLE = \"create_final_covariates\"\n",
"TEXT_UNIT_TABLE = \"create_final_text_units\"\n",
"COMMUNITY_LEVEL = 2"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Read entities"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# read nodes table to get community and degree data\n",
"entity_df = pd.read_parquet(f\"{INPUT_DIR}/{ENTITY_TABLE}.parquet\")\n",
"entity_embedding_df = pd.read_parquet(f\"{INPUT_DIR}/{ENTITY_EMBEDDING_TABLE}.parquet\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Read relationships"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"relationship_df = pd.read_parquet(f\"{INPUT_DIR}/{RELATIONSHIP_TABLE}.parquet\")\n",
"relationships = read_indexer_relationships(relationship_df)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Visualizing nodes and relationships with `yfiles-jupyter-graphs`\n",
"\n",
"`yfiles-jupyter-graphs` is a graph visualization extension that provides interactive and customizable visualizations for structured node and relationship data.\n",
"\n",
"In this case, we use it to provide an interactive visualization for the knowledge graph of the [local_search.ipynb](../../local_search.ipynb) sample by passing node and relationship lists converted from the given parquet files. The requirements for the input data is an `id` attribute for the nodes and `start`/`end` properties for the relationships that correspond to the node ids. Additional attributes can be added in the `properties` of each node/relationship dict:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%pip install yfiles_jupyter_graphs --quiet\n",
"from yfiles_jupyter_graphs import GraphWidget\n",
"\n",
"\n",
"# converts the entities dataframe to a list of dicts for yfiles-jupyter-graphs\n",
"def convert_entities_to_dicts(df):\n",
" \"\"\"Convert the entities dataframe to a list of dicts for yfiles-jupyter-graphs.\"\"\"\n",
" nodes_dict = {}\n",
" for _, row in df.iterrows():\n",
" # Create a dictionary for each row and collect unique nodes\n",
" node_id = row[\"title\"]\n",
" if node_id not in nodes_dict:\n",
" nodes_dict[node_id] = {\n",
" \"id\": node_id,\n",
" \"properties\": row.to_dict(),\n",
" }\n",
" return list(nodes_dict.values())\n",
"\n",
"\n",
"# converts the relationships dataframe to a list of dicts for yfiles-jupyter-graphs\n",
"def convert_relationships_to_dicts(df):\n",
" \"\"\"Convert the relationships dataframe to a list of dicts for yfiles-jupyter-graphs.\"\"\"\n",
" relationships = []\n",
" for _, row in df.iterrows():\n",
" # Create a dictionary for each row\n",
" relationships.append({\n",
" \"start\": row[\"source\"],\n",
" \"end\": row[\"target\"],\n",
" \"properties\": row.to_dict(),\n",
" })\n",
" return relationships\n",
"\n",
"\n",
"w = GraphWidget()\n",
"w.directed = True\n",
"w.nodes = convert_entities_to_dicts(entity_df)\n",
"w.edges = convert_relationships_to_dicts(relationship_df)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure data-driven visualization\n",
"\n",
"The additional properties can be used to configure the visualization for different use cases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# show title on the node\n",
"w.node_label_mapping = \"title\"\n",
"\n",
"\n",
"# map community to a color\n",
"def community_to_color(community):\n",
" \"\"\"Map a community to a color.\"\"\"\n",
" colors = [\n",
" \"crimson\",\n",
" \"darkorange\",\n",
" \"indigo\",\n",
" \"cornflowerblue\",\n",
" \"cyan\",\n",
" \"teal\",\n",
" \"green\",\n",
" ]\n",
" return (\n",
" colors[int(community) % len(colors)] if community is not None else \"lightgray\"\n",
" )\n",
"\n",
"\n",
"def edge_to_source_community(edge):\n",
" \"\"\"Get the community of the source node of an edge.\"\"\"\n",
" source_node = next(\n",
" (entry for entry in w.nodes if entry[\"properties\"][\"title\"] == edge[\"start\"]),\n",
" None,\n",
" )\n",
" source_node_community = source_node[\"properties\"][\"community\"]\n",
" return source_node_community if source_node_community is not None else None\n",
"\n",
"\n",
"w.node_color_mapping = lambda node: community_to_color(node[\"properties\"][\"community\"])\n",
"w.edge_color_mapping = lambda edge: community_to_color(edge_to_source_community(edge))\n",
"# map size data to a reasonable factor\n",
"w.node_scale_factor_mapping = lambda node: 0.5 + node[\"properties\"][\"size\"] * 1.5 / 20\n",
"# use weight for edge thickness\n",
"w.edge_thickness_factor_mapping = \"weight\""
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Automatic layouts\n",
"\n",
"The widget provides different automatic layouts that serve different purposes: `Circular`, `Hierarchic`, `Organic (interactiv or static)`, `Orthogonal`, `Radial`, `Tree`, `Geo-spatial`.\n",
"\n",
"For the knowledge graph, this sample uses the `Circular` layout, though `Hierarchic` or `Organic` are also suitable choices."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Use the circular layout for this visualization. For larger graphs, the default organic layout is often preferrable.\n",
"w.circular_layout()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Display the graph"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"display(w)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Visualizing the result context of `graphrag` queries\n",
"\n",
"The result context of `graphrag` queries allow to inspect the context graph of the request. This data can similarly be visualized as graph with `yfiles-jupyter-graphs`.\n",
"\n",
"## Making the request\n",
"\n",
"The following cell recreates the sample queries from [local_search.ipynb](../../local_search.ipynb)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# setup (see also ../../local_search.ipynb)\n",
"entities = read_indexer_entities(entity_df, entity_embedding_df, COMMUNITY_LEVEL)\n",
"\n",
"description_embedding_store = LanceDBVectorStore(\n",
" collection_name=\"entity_description_embeddings\",\n",
")\n",
"description_embedding_store.connect(db_uri=LANCEDB_URI)\n",
"entity_description_embeddings = store_entity_semantic_embeddings(\n",
" entities=entities, vectorstore=description_embedding_store\n",
")\n",
"covariate_df = pd.read_parquet(f\"{INPUT_DIR}/{COVARIATE_TABLE}.parquet\")\n",
"claims = read_indexer_covariates(covariate_df)\n",
"covariates = {\"claims\": claims}\n",
"report_df = pd.read_parquet(f\"{INPUT_DIR}/{COMMUNITY_REPORT_TABLE}.parquet\")\n",
"reports = read_indexer_reports(report_df, entity_df, COMMUNITY_LEVEL)\n",
"text_unit_df = pd.read_parquet(f\"{INPUT_DIR}/{TEXT_UNIT_TABLE}.parquet\")\n",
"text_units = read_indexer_text_units(text_unit_df)\n",
"\n",
"api_key = os.environ[\"GRAPHRAG_API_KEY\"]\n",
"llm_model = os.environ[\"GRAPHRAG_LLM_MODEL\"]\n",
"embedding_model = os.environ[\"GRAPHRAG_EMBEDDING_MODEL\"]\n",
"\n",
"llm = ChatOpenAI(\n",
" api_key=api_key,\n",
" model=llm_model,\n",
" api_type=OpenaiApiType.OpenAI, # OpenaiApiType.OpenAI or OpenaiApiType.AzureOpenAI\n",
" max_retries=20,\n",
")\n",
"\n",
"token_encoder = tiktoken.get_encoding(\"cl100k_base\")\n",
"\n",
"text_embedder = OpenAIEmbedding(\n",
" api_key=api_key,\n",
" api_base=None,\n",
" api_type=OpenaiApiType.OpenAI,\n",
" model=embedding_model,\n",
" deployment_name=embedding_model,\n",
" max_retries=20,\n",
")\n",
"\n",
"context_builder = LocalSearchMixedContext(\n",
" community_reports=reports,\n",
" text_units=text_units,\n",
" entities=entities,\n",
" relationships=relationships,\n",
" covariates=covariates,\n",
" entity_text_embeddings=description_embedding_store,\n",
" embedding_vectorstore_key=EntityVectorStoreKey.ID, # if the vectorstore uses entity title as ids, set this to EntityVectorStoreKey.TITLE\n",
" text_embedder=text_embedder,\n",
" token_encoder=token_encoder,\n",
")\n",
"\n",
"local_context_params = {\n",
" \"text_unit_prop\": 0.5,\n",
" \"community_prop\": 0.1,\n",
" \"conversation_history_max_turns\": 5,\n",
" \"conversation_history_user_turns_only\": True,\n",
" \"top_k_mapped_entities\": 10,\n",
" \"top_k_relationships\": 10,\n",
" \"include_entity_rank\": True,\n",
" \"include_relationship_weight\": True,\n",
" \"include_community_rank\": False,\n",
" \"return_candidate_context\": False,\n",
" \"embedding_vectorstore_key\": EntityVectorStoreKey.ID, # set this to EntityVectorStoreKey.TITLE if the vectorstore uses entity title as ids\n",
" \"max_tokens\": 12_000, # change this based on the token limit you have on your model (if you are using a model with 8k limit, a good setting could be 5000)\n",
"}\n",
"\n",
"llm_params = {\n",
" \"max_tokens\": 2_000, # change this based on the token limit you have on your model (if you are using a model with 8k limit, a good setting could be 1000=1500)\n",
" \"temperature\": 0.0,\n",
"}\n",
"\n",
"search_engine = LocalSearch(\n",
" llm=llm,\n",
" context_builder=context_builder,\n",
" token_encoder=token_encoder,\n",
" llm_params=llm_params,\n",
" context_builder_params=local_context_params,\n",
" response_type=\"multiple paragraphs\", # free form text describing the response type and format, can be anything, e.g. prioritized list, single paragraph, multiple paragraphs, multiple-page report\n",
")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Run local search on sample queries"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"result = await search_engine.asearch(\"Tell me about Agent Mercer\")\n",
"print(result.response)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"question = \"Tell me about Dr. Jordan Hayes\"\n",
"result = await search_engine.asearch(question)\n",
"print(result.response)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Inspecting the context data used to generate the response"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"result.context_data[\"entities\"].head()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"result.context_data[\"relationships\"].head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Visualizing the result context as graph"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"\"\"\"\n",
"Helper function to visualize the result context with `yfiles-jupyter-graphs`.\n",
"\n",
"The dataframes are converted into supported nodes and relationships lists and then passed to yfiles-jupyter-graphs.\n",
"Additionally, some values are mapped to visualization properties.\n",
"\"\"\"\n",
"\n",
"\n",
"def show_graph(result):\n",
" \"\"\"Visualize the result context with yfiles-jupyter-graphs.\"\"\"\n",
" from yfiles_jupyter_graphs import GraphWidget\n",
"\n",
" if (\n",
" \"entities\" not in result.context_data\n",
" or \"relationships\" not in result.context_data\n",
" ):\n",
" msg = \"The passed results do not contain 'entities' or 'relationships'\"\n",
" raise ValueError(msg)\n",
"\n",
" # converts the entities dataframe to a list of dicts for yfiles-jupyter-graphs\n",
" def convert_entities_to_dicts(df):\n",
" \"\"\"Convert the entities dataframe to a list of dicts for yfiles-jupyter-graphs.\"\"\"\n",
" nodes_dict = {}\n",
" for _, row in df.iterrows():\n",
" # Create a dictionary for each row and collect unique nodes\n",
" node_id = row[\"entity\"]\n",
" if node_id not in nodes_dict:\n",
" nodes_dict[node_id] = {\n",
" \"id\": node_id,\n",
" \"properties\": row.to_dict(),\n",
" }\n",
" return list(nodes_dict.values())\n",
"\n",
" # converts the relationships dataframe to a list of dicts for yfiles-jupyter-graphs\n",
" def convert_relationships_to_dicts(df):\n",
" \"\"\"Convert the relationships dataframe to a list of dicts for yfiles-jupyter-graphs.\"\"\"\n",
" relationships = []\n",
" for _, row in df.iterrows():\n",
" # Create a dictionary for each row\n",
" relationships.append({\n",
" \"start\": row[\"source\"],\n",
" \"end\": row[\"target\"],\n",
" \"properties\": row.to_dict(),\n",
" })\n",
" return relationships\n",
"\n",
" w = GraphWidget()\n",
" # use the converted data to visualize the graph\n",
" w.nodes = convert_entities_to_dicts(result.context_data[\"entities\"])\n",
" w.edges = convert_relationships_to_dicts(result.context_data[\"relationships\"])\n",
" w.directed = True\n",
" # show title on the node\n",
" w.node_label_mapping = \"entity\"\n",
" # use weight for edge thickness\n",
" w.edge_thickness_factor_mapping = \"weight\"\n",
" display(w)\n",
"\n",
"\n",
"show_graph(result)"
]
}
],
"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": "3.10.0"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
|