03_hypergraph.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Hypergraphs\n",
    "\n",
    "> For hypergrapph related functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#| default_exp hypergraph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#| hide\n",
    "from nbdev.showdoc import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# | export\n",
    "import logging\n",
    "import pandas as pd\n",
    "from roux.lib.io import read_dict\n",
    "\n",
    "import roux.lib.df as rd #noqa\n",
    "\n",
    "def get_degreeby_subset(\n",
    "    nodes: pd.DataFrame,\n",
    "    edges: pd.DataFrame,\n",
    "    col_node_id: str,\n",
    "    col_source: str,\n",
    "    col_target: str,\n",
    "    col_subset_id: str,\n",
    ") -> pd.DataFrame:\n",
    "    \"\"\"Get degrees by subsets (groups) of the source nodes\n",
    "\n",
    "    Args:\n",
    "        nodes (pd.DataFrame): nodes\n",
    "        edges (pd.DataFrame): edges\n",
    "        col_node_id (str): column with node ids\n",
    "        col_source (str): column with source ids\n",
    "        col_target (str): column with target ids\n",
    "        col_subset_id (str): column with subset ids\n",
    "\n",
    "    Returns:\n",
    "        pd.DataFrame: table\n",
    "    \"\"\"\n",
    "    import networkx as nx\n",
    "\n",
    "    G = nx.from_pandas_edgelist(edges, col_source, col_target)\n",
    "\n",
    "    return (\n",
    "        nodes.loc[:, [col_node_id]]\n",
    "        .rename(columns={col_node_id: \"source\"})\n",
    "        .assign(\n",
    "            **{\n",
    "                \"target\": lambda df: df[\"source\"].apply(\n",
    "                    lambda x: list(G.neighbors(x)) if x in G.nodes else None\n",
    "                )\n",
    "            }\n",
    "        )\n",
    "        .explode(\"target\")\n",
    "        .log.dropna()\n",
    "        .merge(\n",
    "            right=nodes.loc[:, [col_node_id, col_subset_id]].rename(\n",
    "                columns={col_node_id: \"target\"}\n",
    "            ),\n",
    "            on=\"target\",\n",
    "            how=\"left\",\n",
    "        )\n",
    "        .groupby(\"source\")[col_subset_id]\n",
    "        .agg(lambda x: x.value_counts(normalize=True).to_dict())\n",
    "        .apply(pd.Series)\n",
    "        .fillna(0)\n",
    "        .reset_index()\n",
    "    )\n",
    "\n",
    "\n",
    "def plot_degreeby_subset(\n",
    "    nodes: pd.DataFrame,\n",
    "    edges: pd.DataFrame,\n",
    "    col_node_id: str,\n",
    "    col_source: str,\n",
    "    col_target: str,\n",
    "    col_subset_id: str,\n",
    "    order: list = None,\n",
    "    hue_order: list = None,\n",
    "    config_base_path: str = None,\n",
    "    cmap_subsets: str = None,\n",
    "    show_text: bool = True,\n",
    "    plot: bool = True,\n",
    "    **kws_display_plot,\n",
    ") -> tuple:\n",
    "    \"\"\"Plot bar plot of degrees by subsets\n",
    "\n",
    "    Args:\n",
    "        nodes (pd.DataFrame): nodes\n",
    "        edges (pd.DataFrame): edges\n",
    "        col_node_id (str): column with node ids\n",
    "        col_source (str): column with sources\n",
    "        col_target (str): column with targets\n",
    "        col_subset_id (str): column with subset ids\n",
    "        order (list, optional): order of subsets. Defaults to None.\n",
    "        hue_order (list, optional): order of colors. Defaults to None.\n",
    "        config_base_path (str, optional): base vega config. Defaults to None.\n",
    "        cmap_subsets (str, optional): colormap. Defaults to None.\n",
    "        show_text (bool, optional): show text. Defaults to True.\n",
    "        plot (bool, optional): plot or not. Defaults to True.\n",
    "\n",
    "    Returns:\n",
    "        tuple: vega config, degrees\n",
    "    \"\"\"\n",
    "    degreebysubset = get_degreeby_subset(\n",
    "        nodes,\n",
    "        edges,\n",
    "        col_node_id=col_node_id,\n",
    "        col_source=col_source,\n",
    "        col_target=col_target,\n",
    "        col_subset_id=col_subset_id,\n",
    "    )\n",
    "    if not order is None:\n",
    "        degreebysubset = degreebysubset.set_index(\"source\").loc[order, :].reset_index()\n",
    "    # degreebysubset.head(1)\n",
    "\n",
    "    if config_base_path is None:\n",
    "        from parag.utils import get_src_path\n",
    "\n",
    "        config_base_path = f\"{get_src_path()}/data/stacked_bars.json\"\n",
    "    cfg_bars = read_dict(config_base_path)\n",
    "\n",
    "    ## data\n",
    "    cfg_bars[\"data\"][0][\"values\"] = list(degreebysubset.T.to_dict().values())\n",
    "\n",
    "    # if cmap_subsets is None:\n",
    "    #     import seaborn as sns # for the matching color palette\n",
    "    #     cmap_subsets=sns.color_palette(\"colorblind\")\n",
    "\n",
    "    from roux.viz.colors import get_ncolors\n",
    "\n",
    "    if hue_order is None:\n",
    "        subsets = degreebysubset.set_index(\"source\").columns.tolist()\n",
    "    else:\n",
    "        subsets = hue_order\n",
    "    from parag.utils import get_colors\n",
    "\n",
    "    to_subset_color = get_colors(\n",
    "        subsets,\n",
    "        cmap_subsets=cmap_subsets,\n",
    "    )\n",
    "    # to_subset_color=dict(zip(subsets,get_ncolors(len(subsets),cmap=cmap_subsets)))\n",
    "\n",
    "    ## \"fields\"\n",
    "    cfg_bars[\"data\"][0][\"transform\"][0][\"fields\"] = degreebysubset.set_index(\n",
    "        \"source\"\n",
    "    ).columns.tolist()\n",
    "\n",
    "    ## formula\n",
    "    expr = (\n",
    "        \",\".join([f\"if(datum.key==='{f}','{c}'\" for f, c in to_subset_color.items()])\n",
    "    ) + \",'#fff'\"\n",
    "    ## close brackets\n",
    "    expr += \")\" * (expr.count(\"(\") - expr.count(\")\"))\n",
    "\n",
    "    cfg_bars[\"data\"][0][\"transform\"][1][\"expr\"] = expr\n",
    "\n",
    "    # \"if(datum.key==='CD4+/CD45RO+ Memory','green',if(datum.key==='CD4+/CD25 T Reg','#f0f','red'))\"\n",
    "\n",
    "    ## todo set the text\n",
    "    if not show_text:\n",
    "        cfg_bars[\"marks\"] = cfg_bars[\"marks\"][:1]\n",
    "    if plot:\n",
    "        from parag.utils import display_plot\n",
    "\n",
    "        display_plot(cfg_bars.copy(), **kws_display_plot)\n",
    "    return cfg_bars, degreebysubset\n",
    "\n",
    "\n",
    "# parameters\n",
    "def to_net(\n",
    "    nodes: pd.DataFrame,\n",
    "    edges: pd.DataFrame,\n",
    "    col_node_id: str,\n",
    "    col_source: str,\n",
    "    col_target: str,\n",
    "    col_subset_id: str,\n",
    "    show_node_names: bool = True,\n",
    "    defaults: dict = None,\n",
    "    dbug: bool = False,\n",
    ") -> tuple:\n",
    "    \"\"\"Plot interactive hypergraph 'representation'.\n",
    "\n",
    "    Args:\n",
    "        nodes (pd.DataFrame): nodes\n",
    "        edges (pd.DataFrame): edges\n",
    "        col_node_id (str): column with node ids\n",
    "        col_source (str): column with sources\n",
    "        col_target (str): column with targets\n",
    "        col_subset_id (str): column with subset ids\n",
    "        show_node_names (bool, optional): show node names. Defaults to True.\n",
    "        defaults (dict, optional): defaults provided to the vega config. Defaults to None.\n",
    "        dbug (bool, optional): debug mode. Defaults to False.\n",
    "\n",
    "    Returns:\n",
    "        tuple: vega config, degrees\n",
    "    \"\"\"\n",
    "    from parag.graph import to_net\n",
    "\n",
    "    # %run parag/graph.py\n",
    "    cfg_base, df1 = to_net(\n",
    "        nodes=nodes,\n",
    "        edges=edges,\n",
    "        col_node_id=col_node_id,\n",
    "        col_source=col_source,\n",
    "        col_target=col_target,\n",
    "        col_subset_id=col_subset_id,\n",
    "        plot=dbug,\n",
    "        defaults=dict(\n",
    "            radius=165,\n",
    "            textSize=8,\n",
    "            textOffset=7,\n",
    "        ),\n",
    "    )\n",
    "\n",
    "    ## sorted bars\n",
    "    cfg_bars, degreebysubset = plot_degreeby_subset(\n",
    "        nodes,\n",
    "        edges,\n",
    "        col_node_id=col_node_id,\n",
    "        col_source=col_source,\n",
    "        col_target=col_target,\n",
    "        col_subset_id=col_subset_id,\n",
    "        config_base_path=None,\n",
    "        cmap_subsets=None,\n",
    "        plot=dbug,\n",
    "        order=df1.log.query(expr=\"`parent`==3\")[\"name\"].tolist(),\n",
    "        hue_order=df1[col_subset_id].dropna().unique(),\n",
    "    )\n",
    "\n",
    "    cfg_hyg = dict(zip(cfg_base.keys(), cfg_base.values()))\n",
    "\n",
    "    cfg_hyg[\"scales\"] = cfg_bars[\"scales\"]\n",
    "\n",
    "    cfg_hyg[\"signals\"] += cfg_bars[\"signals\"]\n",
    "    cfg_hyg[\"data\"] += cfg_bars[\"data\"]\n",
    "    cfg_hyg[\"marks\"] += cfg_bars[\"marks\"]\n",
    "\n",
    "    ## removal of unsupported signals\n",
    "    for i,d in enumerate(cfg_hyg['signals']):\n",
    "        if d['name'] in ['rotate','startAngle','endAngle','extent',\"layout\"]:\n",
    "            del cfg_hyg['signals'][i]['bind']\n",
    "    \n",
    "    ## aes\n",
    "    if not show_node_names:\n",
    "        for k in [\"opacity\", \"fillOpacity\", \"strokeOpacity\"]:\n",
    "            cfg_hyg[\"marks\"][0][\"encode\"][\"enter\"][k] = {\"value\": 0}\n",
    "            cfg_hyg[\"marks\"][0][\"encode\"][\"update\"][k] = {\"value\": 0}\n",
    "\n",
    "    from parag.utils import display_plot\n",
    "    \n",
    "    try:\n",
    "        display_plot(\n",
    "            cfg_hyg.copy(),\n",
    "            defaults=defaults,\n",
    "        )\n",
    "    except DisplayError:\n",
    "        logging.error(\"plot not displayed, check the config.\")\n",
    "    return cfg_hyg, degreebysubset\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Demo data**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from parag.core import get_net_data\n",
    "import scanpy as sc\n",
    "nodes,edges=get_net_data(sc.datasets.pbmc68k_reduced())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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>cell id</th>\n",
       "      <th>bulk_labels</th>\n",
       "      <th>n_genes</th>\n",
       "      <th>percent_mito</th>\n",
       "      <th>n_counts</th>\n",
       "      <th>S_score</th>\n",
       "      <th>G2M_score</th>\n",
       "      <th>phase</th>\n",
       "      <th>louvain</th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CAGACAACAAAACG-7</td>\n",
       "      <td>CD4+/CD25 T Reg</td>\n",
       "      <td>1109</td>\n",
       "      <td>0.012702</td>\n",
       "      <td>3779.0</td>\n",
       "      <td>-0.054589</td>\n",
       "      <td>-1.267457</td>\n",
       "      <td>G1</td>\n",
       "      <td>0</td>\n",
       "      <td>-7.174749</td>\n",
       "      <td>-6.994063</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            cell id      bulk_labels  n_genes  percent_mito  n_counts  \\\n",
       "0  CAGACAACAAAACG-7  CD4+/CD25 T Reg     1109      0.012702    3779.0   \n",
       "\n",
       "    S_score  G2M_score phase louvain         x         y  \n",
       "0 -0.054589  -1.267457    G1       0 -7.174749 -6.994063  "
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nodes.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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>cell id1</th>\n",
       "      <th>cell id2</th>\n",
       "      <th>distance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CAGACAACAAAACG-7</td>\n",
       "      <td>GTGATGACTGGGAG-4</td>\n",
       "      <td>1.298669</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           cell id1          cell id2  distance\n",
       "0  CAGACAACAAAACG-7  GTGATGACTGGGAG-4  1.298669"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "edges.head(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Degrees by subsets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "cfg_bars=plot_degreeby_subset(\n",
    "    nodes.sort_values('bulk_labels',ascending=True),\n",
    "    edges,\n",
    "    col_node_id='cell id',\n",
    "    col_source='cell id1',\n",
    "    col_target='cell id2',\n",
    "    col_subset_id='bulk_labels',    \n",
    "    config_base_path=None,\n",
    "    cmap_subsets=None,  \n",
    "    # defaults=dict(\n",
    "    #     innerRadius=170,\n",
    "    #     outerRadius=200,\n",
    "    #     innerPadding=0.15,\n",
    "    # ),\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Integrate with graph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "cfg,df_=to_net(\n",
    "    nodes,\n",
    "    edges,\n",
    "    col_node_id='cell id',\n",
    "    col_source='cell id1',\n",
    "    col_target='cell id2',\n",
    "    col_subset_id='bulk_labels',\n",
    "    show_node_names=False,\n",
    "    defaults=dict(\n",
    "            textSize=8,\n",
    "            textOffset=3,\n",
    "    ),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#| hide\n",
    "import nbdev; nbdev.nbdev_export()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "parag",
   "language": "python",
   "name": "parag"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}