{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "f2a11223",
   "metadata": {},
   "source": [
    "# Exploring NAT4 \u2014 the deepSTRF data paradigm in action\n",
    "\n",
    "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/urancon/deepSTRF/blob/develop/examples/explore_nat4.ipynb)\n",
    "\n",
    "This notebook walks through the **NAT4** dataset (Pennington & David\n",
    "2023; auto-downloaded from Zenodo) as a hands-on tour of the deepSTRF\n",
    "data API. NAT4 is the dataset that exercises every subtle bit of the\n",
    "[`data_paradigm.md`](../docs/_source/md/data_paradigm.md):\n",
    "\n",
    "- **Sparse coverage** \u2014 not every cell saw every stimulus (33 of 849 A1\n",
    "  cells have no validation data); missing pairs are NaN-padded sentinels.\n",
    "- **Estimation / validation split** \u2014 575 est + 18 val stims, with a\n",
    "  `stim_meta['subset']` tag and a `subset='all'|'est'|'val'`\n",
    "  constructor shortcut.\n",
    "- **Per-cell metadata** richness \u2014 cell IDs parsed into site / animal /\n",
    "  electrode / unit-in-electrode + auditory / depth tags.\n",
    "- **The bidirectional rule** \u2014 selecting a stim subset auto-hides\n",
    "  neurons whose only valid responses lie outside that subset.\n",
    "- **`valid_mask`** \u2014 the bool grid the dataloader emits alongside\n",
    "  responses, derived from NaN sentinels at collate time.\n",
    "\n",
    "We focus on A1 here; the notebook closes with a brief PEG comparison and\n",
    "a cross-area concatenation example.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7d3aca46",
   "metadata": {},
   "source": [
    "## Setup \u2014 Google Colab\n",
    "\n",
    "If you're running on Google Colab, the cell below installs deepSTRF\n",
    "from source. On a local install (`pip install -e .`) it's a no-op."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ebc31ee0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "if 'google.colab' in sys.modules:\n",
    "    !pip install -q git+https://github.com/urancon/deepSTRF.git\n",
    "    print(\"deepSTRF installed from GitHub.\")\n",
    "else:\n",
    "    print(\"Local environment \u2014 assuming deepSTRF is already importable.\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6b0a241f",
   "metadata": {},
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c2afc914",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-02T16:47:42.109871Z",
     "iopub.status.busy": "2026-05-02T16:47:42.109683Z",
     "iopub.status.idle": "2026-05-02T16:47:44.456200Z",
     "shell.execute_reply": "2026-05-02T16:47:44.455404Z"
    }
   },
   "outputs": [],
   "source": "%matplotlib inline\nimport matplotlib.pyplot as plt\nimport numpy as np\nimport torch\nfrom collections import Counter\nfrom torch.utils.data import DataLoader, Subset\n\nfrom deepSTRF.datasets.audio.nat4 import NAT4Dataset\nfrom deepSTRF.utils import (\n    neural_collate, concat_neural_datasets, plot_stim_with_response,\n)"
  },
  {
   "cell_type": "markdown",
   "id": "09598c37",
   "metadata": {},
   "source": [
    "## 1. Load NAT4 A1\n",
    "\n",
    "`NAT4Dataset(area='A1', download=True)` fetches 3 archives from Zenodo\n",
    "(~30 MB main + ~70 MB single-sites + 70 kB CSV) and unpacks them under\n",
    "`~/.cache/deepSTRF/NAT4/A1/`. Subsequent loads use the cache. The\n",
    "default `subset='all'` loads both estimation and validation stims; for\n",
    "training-only workflows, `subset='est'` loads only the 575 estimation\n",
    "stims (faster)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4b44814e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-02T16:47:44.458247Z",
     "iopub.status.busy": "2026-05-02T16:47:44.458013Z",
     "iopub.status.idle": "2026-05-02T16:48:12.315078Z",
     "shell.execute_reply": "2026-05-02T16:48:12.314459Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:   0%|          | 0/22 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:   5%|\u258d         | 1/22 [00:00<00:07,  2.84it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:   9%|\u2589         | 2/22 [00:00<00:05,  3.78it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  14%|\u2588\u258e        | 3/22 [00:01<00:06,  2.87it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  18%|\u2588\u258a        | 4/22 [00:01<00:06,  2.70it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  23%|\u2588\u2588\u258e       | 5/22 [00:01<00:05,  3.32it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  27%|\u2588\u2588\u258b       | 6/22 [00:01<00:04,  3.39it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  32%|\u2588\u2588\u2588\u258f      | 7/22 [00:02<00:03,  4.09it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  36%|\u2588\u2588\u2588\u258b      | 8/22 [00:02<00:06,  2.14it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  41%|\u2588\u2588\u2588\u2588      | 9/22 [00:03<00:06,  2.01it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  45%|\u2588\u2588\u2588\u2588\u258c     | 10/22 [00:04<00:06,  1.78it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  50%|\u2588\u2588\u2588\u2588\u2588     | 11/22 [00:04<00:05,  1.98it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  55%|\u2588\u2588\u2588\u2588\u2588\u258d    | 12/22 [00:04<00:04,  2.39it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  59%|\u2588\u2588\u2588\u2588\u2588\u2589    | 13/22 [00:05<00:03,  2.63it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  64%|\u2588\u2588\u2588\u2588\u2588\u2588\u258e   | 14/22 [00:05<00:02,  3.19it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  68%|\u2588\u2588\u2588\u2588\u2588\u2588\u258a   | 15/22 [00:05<00:02,  3.50it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  73%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u258e  | 16/22 [00:06<00:02,  2.81it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  77%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u258b  | 17/22 [00:06<00:01,  2.69it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  82%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u258f | 18/22 [00:06<00:01,  2.57it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  86%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u258b | 19/22 [00:07<00:01,  2.35it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  91%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588 | 20/22 [00:08<00:01,  1.99it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites:  95%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u258c| 21/22 [00:08<00:00,  1.99it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites: 100%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588| 22/22 [00:08<00:00,  2.31it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 A1 val sites: 100%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588| 22/22 [00:08<00:00,  2.50it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cells:    849\n",
      "Stimuli:  593\n",
      "  est:    575\n",
      "  val:    18\n",
      "Spec shape: (1, 18, 150)  (1, F, T) at dt=10.0 ms\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "ds = NAT4Dataset(area='A1', download=True, subset='all')\n",
    "print(f\"Cells:    {ds.N_neurons}\")\n",
    "print(f\"Stimuli:  {len(ds.stims)}\")\n",
    "counts = Counter(m['subset'] for m in ds.stim_meta)\n",
    "print(f\"  est:    {counts['est']}\")\n",
    "print(f\"  val:    {counts['val']}\")\n",
    "print(f\"Spec shape: {tuple(ds.stims[0].shape)}  (1, F, T) at dt={ds.dt} ms\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2266a8c1",
   "metadata": {},
   "source": [
    "## 2. Per-cell metadata\n",
    "\n",
    "Every neuron carries a metadata dict with the cell ID parsed into its\n",
    "recording site, animal (3-letter prefix), electrode, and\n",
    "unit-in-electrode. NAT4 also marks the auditory-vs-non-auditory\n",
    "classification used by Pennington & David (2023)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "b854a600",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-02T16:48:12.316973Z",
     "iopub.status.busy": "2026-05-02T16:48:12.316831Z",
     "iopub.status.idle": "2026-05-02T16:48:12.320723Z",
     "shell.execute_reply": "2026-05-02T16:48:12.320099Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "First 3 entries:\n",
      "  {'cell_id': 'ARM029a-04-1', 'area': 'A1', 'auditory': True, 'site': 'ARM029a', 'animal': 'ARM', 'electrode': 4, 'unit_in_electrode': 1}\n",
      "  {'cell_id': 'ARM029a-07-6', 'area': 'A1', 'auditory': True, 'site': 'ARM029a', 'animal': 'ARM', 'electrode': 7, 'unit_in_electrode': 6}\n",
      "  {'cell_id': 'ARM029a-07-7', 'area': 'A1', 'auditory': True, 'site': 'ARM029a', 'animal': 'ARM', 'electrode': 7, 'unit_in_electrode': 7}\n",
      "\n",
      "auditory cells: 777 / 849\n",
      "animals:        {'ARM': 165, 'CRD': 36, 'DRX': 244, 'TNC': 404}\n",
      "recording sites: 22 unique  (top 5: [('DRX006b', 95), ('DRX008b', 86), ('DRX007a', 63), ('TNC018a', 60), ('TNC017a', 57)])\n"
     ]
    }
   ],
   "source": [
    "print(\"First 3 entries:\")\n",
    "for n in ds.nrn_meta[:3]:\n",
    "    print(f\"  {n}\")\n",
    "print()\n",
    "n_auditory = sum(1 for n in ds.nrn_meta if n['auditory'])\n",
    "print(f\"auditory cells: {n_auditory} / {ds.N_neurons}\")\n",
    "animals = Counter(n['animal'] for n in ds.nrn_meta)\n",
    "print(f\"animals:        {dict(animals)}\")\n",
    "sites = Counter(n['site'] for n in ds.nrn_meta)\n",
    "print(f\"recording sites: {len(sites)} unique  (top 5: {sites.most_common(5)})\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bfdb456c",
   "metadata": {},
   "source": [
    "## 3. Coverage map \u2014 `nrn_masks`\n",
    "\n",
    "`ds.nrn_masks` is a `(S, N)` bool tensor: `True` where neuron `n` has\n",
    "recorded responses for stim `s`. NAT4 has dense est coverage but sparse\n",
    "val coverage \u2014 the visualisation below shows the block structure."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e005c1c1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-02T16:48:12.322130Z",
     "iopub.status.busy": "2026-05-02T16:48:12.321997Z",
     "iopub.status.idle": "2026-05-02T16:48:14.669359Z",
     "shell.execute_reply": "2026-05-02T16:48:14.668357Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "valid (stim, cell) pairs: 502,863 / 503,457  (99.9%)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cells with \u22651 val response: 816 / 849\n",
      "cells with NO val data:     33\n"
     ]
    }
   ],
   "source": [
    "masks = ds.nrn_masks                                      # (S=593, N=849)\n",
    "S, N = masks.shape\n",
    "total = S * N\n",
    "valid = masks.sum().item()\n",
    "print(f\"valid (stim, cell) pairs: {valid:,} / {total:,}  ({100*valid/total:.1f}%)\")\n",
    "\n",
    "# Order stims by est/val and cells by val-data presence so the structure is visible.\n",
    "stim_order = sorted(range(S), key=lambda s: ds.stim_meta[s]['subset'])\n",
    "val_idx = [s for s in stim_order if ds.stim_meta[s]['subset'] == 'val']\n",
    "n_val_stims = len(val_idx)\n",
    "\n",
    "has_val = masks[val_idx].any(dim=0)                       # (N,) \u2014 True if cell has any val response\n",
    "cell_order = torch.argsort(has_val.long(), descending=True)\n",
    "\n",
    "reordered = masks[stim_order][:, cell_order]\n",
    "fig, ax = plt.subplots(figsize=(9, 4))\n",
    "ax.imshow(reordered.numpy().T, aspect='auto', cmap='Greys', interpolation='nearest')\n",
    "ax.axvline(S - n_val_stims - 0.5, color='red', lw=1.5, label=f'est | val boundary')\n",
    "ax.set_xlabel(\"stim (est on left, val on right)\")\n",
    "ax.set_ylabel(\"cell (cells with val data on top)\")\n",
    "ax.set_title(f\"NAT4 A1 nrn_masks \u2014 {valid:,} / {total:,} valid pairs\")\n",
    "ax.legend(loc='upper right')\n",
    "plt.tight_layout(); plt.show()\n",
    "\n",
    "n_cells_with_val = has_val.sum().item()\n",
    "print(f\"cells with \u22651 val response: {n_cells_with_val} / {N}\")\n",
    "print(f\"cells with NO val data:     {N - n_cells_with_val}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f652e1d1",
   "metadata": {},
   "source": [
    "## 4. Stim metadata + one example stim/response\n",
    "\n",
    "The estimation stims and validation stims share the same\n",
    "spectrogram-extraction pipeline; what differs is the trial count\n",
    "(typically 1 repeat for est vs ~10-20 for val). Below: a validation\n",
    "stim spectrogram and the response of one auditory cell."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "29f5056a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-02T16:48:14.670896Z",
     "iopub.status.busy": "2026-05-02T16:48:14.670761Z",
     "iopub.status.idle": "2026-05-02T16:48:14.914432Z",
     "shell.execute_reply": "2026-05-02T16:48:14.913831Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x550 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": "# Find a val stim with a high-coverage auditory cell.\nval_stim_indices = [s for s, m in enumerate(ds.stim_meta) if m['subset'] == 'val']\nauditory_cell_indices = [i for i, n in enumerate(ds.nrn_meta) if n['auditory']]\nfor cell_idx in auditory_cell_indices:\n    valid_val_stims = [s for s in val_stim_indices if not ds.responses[s][cell_idx].isnan().all()]\n    if valid_val_stims:\n        stim_idx = valid_val_stims[0]\n        break\n\nplot_stim_with_response(\n    ds.stims[stim_idx], ds.responses[stim_idx][cell_idx], dt_ms=ds.dt,\n    title=(f\"NAT4 A1 stim {stim_idx} ('{ds.stim_meta[stim_idx]['name']}', \"\n           f\"subset={ds.stim_meta[stim_idx]['subset']}) \"\n           f\"\u2192 cell {cell_idx} ({ds.nrn_meta[cell_idx]['cell_id']})\"),\n)\nplt.show()"
  },
  {
   "cell_type": "markdown",
   "id": "80336a49",
   "metadata": {},
   "source": [
    "## 5. The bidirectional rule\n",
    "\n",
    "`select_stims_by_attr('subset', 'val')` filters down to validation\n",
    "stims. Because we now restrict the stim set, neurons whose **only**\n",
    "valid responses lie outside the selection (i.e. the val-less cells\n",
    "visible in \u00a73) are auto-hidden. This is the bidirectional rule from\n",
    "[`data_paradigm.md`](../docs/_source/md/data_paradigm.md) \u00a78: a stim\n",
    "selection implicitly induces a population selection.\n",
    "\n",
    "After the selection:\n",
    "- Stims drop from 593 to 18 (only validation stims remain).\n",
    "- Cells drop from 849 to 816 (the 33 val-less cells are hidden)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d387df2f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-02T16:48:14.916365Z",
     "iopub.status.busy": "2026-05-02T16:48:14.916221Z",
     "iopub.status.idle": "2026-05-02T16:48:27.907738Z",
     "shell.execute_reply": "2026-05-02T16:48:27.906947Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "after select_stims_by_attr('subset', 'val'):\n",
      "  visible stims:  18\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  visible cells:  816\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "first returned item:\n",
      "  stim shape:  (1, 18, 150)\n",
      "  N returned:  816\n",
      "  meta:        {'name': 'STIM_00cat172_rec1_geese_excerpt1.wav', 'subset': 'val'}\n"
     ]
    }
   ],
   "source": [
    "ds.select_stims_by_attr('subset', 'val')\n",
    "\n",
    "n_visible_cells = ds.nrn_masks.sum(dim=0).gt(0).sum().item() if ds.S_sel is None else len(ds._selected())\n",
    "print(f\"after select_stims_by_attr('subset', 'val'):\")\n",
    "print(f\"  visible stims:  {len(ds.S_sel) if ds.S_sel is not None else len(ds.stims)}\")\n",
    "print(f\"  visible cells:  {len(ds._selected())}\")\n",
    "\n",
    "# What __getitem__ returns now: only val-data cells, only val stims.\n",
    "item = ds[0]\nstim, responses_per_neuron, meta = item['stims'], item['responses'], item['stim_meta']\n",
    "print(f\"\\nfirst returned item:\")\n",
    "print(f\"  stim shape:  {tuple(stim.shape)}\")\n",
    "print(f\"  N returned:  {len(responses_per_neuron)}\")\n",
    "print(f\"  meta:        {meta}\")\n",
    "\n",
    "# undo selection so subsequent cells see the full dataset\n",
    "ds.reset_stim_selection()\n",
    "ds.reset_pop_selection()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4380ec98",
   "metadata": {},
   "source": [
    "## 6. Population filtering \u2014 `select_pop_by_nrn_attr`\n",
    "\n",
    "The neuron-axis filters work on any key in `nrn_meta`. Two\n",
    "common cases: keeping only auditory cells, and slicing by animal /\n",
    "recording site."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "fe318615",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-02T16:48:27.909185Z",
     "iopub.status.busy": "2026-05-02T16:48:27.909046Z",
     "iopub.status.idle": "2026-05-02T16:48:27.912504Z",
     "shell.execute_reply": "2026-05-02T16:48:27.911684Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "after select_pop_by_nrn_attr('auditory', True): 777 cells\n",
      "after select_pop_by_nrn_attr('animal', 'ARM'):  165 cells\n"
     ]
    }
   ],
   "source": [
    "ds.select_pop_by_nrn_attr('auditory', True)\n",
    "print(f\"after select_pop_by_nrn_attr('auditory', True): {len(ds._selected())} cells\")\n",
    "ds.reset_pop_selection()\n",
    "\n",
    "ds.select_pop_by_nrn_attr('animal', 'ARM')\n",
    "print(f\"after select_pop_by_nrn_attr('animal', 'ARM'):  {len(ds._selected())} cells\")\n",
    "ds.reset_pop_selection()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "be3bdd42",
   "metadata": {},
   "source": [
    "## 7. The `valid_mask` from `neural_collate`\n",
    "\n",
    "When the dataset feeds a `DataLoader` with `neural_collate`, the\n",
    "collator emits a fine-grained `(B, N, R, T)` bool mask alongside the\n",
    "NaN-padded responses tensor. Loss code uses this directly (see\n",
    "[`metrics_paradigm.md`](../docs/_source/md/metrics_paradigm.md) \u00a74) \u2014\n",
    "no need to re-scan for NaN positions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "f702acbb",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-02T16:48:27.914082Z",
     "iopub.status.busy": "2026-05-02T16:48:27.913946Z",
     "iopub.status.idle": "2026-05-02T16:49:07.403599Z",
     "shell.execute_reply": "2026-05-02T16:49:07.403001Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "stims:      (4, 1, 18, 150)  (B, 1, F, T)\n",
      "responses:  (4, 744, 20, 150)  (B, N, R_max, T_max)\n",
      "valid_mask: (4, 744, 20, 150)  bool\n",
      "NaN cells in responses: 0 / 8,928,000\n",
      "valid_mask True count:  8,928,000  (should equal numel - NaN count)\n"
     ]
    }
   ],
   "source": [
    "# small batched loader (val stims, restricted to auditory cells)\n",
    "ds.select_stims_by_attr('subset', 'val')\n",
    "ds.select_pop_by_nrn_attr('auditory', True)\n",
    "loader = DataLoader(ds, batch_size=4, shuffle=False, collate_fn=neural_collate)\n",
    "batch = next(iter(loader))\nstims, responses, valid_mask, stim_metas = batch['stims'], batch['responses'], batch['valid_mask'], batch['stim_meta']\n",
    "print(f\"stims:      {tuple(stims.shape)}  (B, 1, F, T)\")\n",
    "print(f\"responses:  {tuple(responses.shape)}  (B, N, R_max, T_max)\")\n",
    "print(f\"valid_mask: {tuple(valid_mask.shape)}  bool\")\n",
    "print(f\"NaN cells in responses: {responses.isnan().sum().item():,} / {responses.numel():,}\")\n",
    "print(f\"valid_mask True count:  {valid_mask.sum().item():,}  (should equal numel - NaN count)\")\n",
    "ds.reset_stim_selection()\n",
    "ds.reset_pop_selection()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b0902da",
   "metadata": {},
   "source": [
    "## 8. The PEG companion\n",
    "\n",
    "NAT4 ships a second area, **PEG** (parabelt secondary auditory cortex).\n",
    "Same protocol, different population."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "fb415d70",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-02T16:49:07.405282Z",
     "iopub.status.busy": "2026-05-02T16:49:07.405132Z",
     "iopub.status.idle": "2026-05-02T16:49:27.969288Z",
     "shell.execute_reply": "2026-05-02T16:49:27.968576Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ulysse/miniconda3/envs/deepstrf_dev/lib/python3.11/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_NAT4_ozgf.fs100.ch18.tgz:   0%|          | 0.00/22.4M [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_NAT4_ozgf.fs100.ch18.tgz:   5%|\u258d         | 1.05M/22.4M [00:00<00:12, 1.71MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_NAT4_ozgf.fs100.ch18.tgz:   9%|\u2589         | 2.10M/22.4M [00:00<00:07, 2.80MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_NAT4_ozgf.fs100.ch18.tgz:  19%|\u2588\u2589        | 4.19M/22.4M [00:00<00:03, 5.96MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_NAT4_ozgf.fs100.ch18.tgz:  38%|\u2588\u2588\u2588\u258a      | 8.39M/22.4M [00:01<00:01, 12.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_NAT4_ozgf.fs100.ch18.tgz:  56%|\u2588\u2588\u2588\u2588\u2588\u258b    | 12.6M/22.4M [00:01<00:00, 16.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_NAT4_ozgf.fs100.ch18.tgz:  66%|\u2588\u2588\u2588\u2588\u2588\u2588\u258c   | 14.7M/22.4M [00:01<00:00, 14.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_NAT4_ozgf.fs100.ch18.tgz:  75%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u258c  | 16.8M/22.4M [00:01<00:00, 13.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_NAT4_ozgf.fs100.ch18.tgz:  84%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u258d | 18.9M/22.4M [00:01<00:00, 12.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_NAT4_ozgf.fs100.ch18.tgz:  94%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u258d| 21.0M/22.4M [00:02<00:00, 12.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_NAT4_ozgf.fs100.ch18.tgz: 100%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588| 22.4M/22.4M [00:02<00:00, 11.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_NAT4_ozgf.fs100.ch18.tgz: 100%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588| 22.4M/22.4M [00:02<00:00, 10.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_pred_correlation.csv: 0.00B [00:00, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_pred_correlation.csv: 33.0kB [00:00, 959kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_single_sites.zip:   0%|          | 0.00/25.4M [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_single_sites.zip:   4%|\u258d         | 1.05M/25.4M [00:00<00:18, 1.32MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_single_sites.zip:  12%|\u2588\u258f        | 3.15M/25.4M [00:00<00:05, 4.21MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_single_sites.zip:  29%|\u2588\u2588\u2589       | 7.34M/25.4M [00:01<00:01, 10.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_single_sites.zip:  45%|\u2588\u2588\u2588\u2588\u258c     | 11.5M/25.4M [00:01<00:00, 15.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_single_sites.zip:  62%|\u2588\u2588\u2588\u2588\u2588\u2588\u258f   | 15.7M/25.4M [00:01<00:00, 20.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_single_sites.zip:  74%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u258d  | 18.9M/25.4M [00:01<00:00, 22.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_single_sites.zip:  87%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u258b | 22.0M/25.4M [00:01<00:00, 19.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_single_sites.zip:  99%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2589| 25.2M/25.4M [00:01<00:00, 14.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "download PEG_single_sites.zip: 100%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588| 25.4M/25.4M [00:02<00:00, 12.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 PEG val sites:   0%|          | 0/12 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 PEG val sites:   8%|\u258a         | 1/12 [00:00<00:05,  1.90it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 PEG val sites:  17%|\u2588\u258b        | 2/12 [00:00<00:03,  2.59it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 PEG val sites:  25%|\u2588\u2588\u258c       | 3/12 [00:01<00:04,  2.03it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 PEG val sites:  33%|\u2588\u2588\u2588\u258e      | 4/12 [00:01<00:03,  2.05it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 PEG val sites:  42%|\u2588\u2588\u2588\u2588\u258f     | 5/12 [00:02<00:02,  2.42it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 PEG val sites:  58%|\u2588\u2588\u2588\u2588\u2588\u258a    | 7/12 [00:02<00:01,  3.10it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 PEG val sites:  67%|\u2588\u2588\u2588\u2588\u2588\u2588\u258b   | 8/12 [00:02<00:01,  3.38it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 PEG val sites:  75%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u258c  | 9/12 [00:03<00:00,  3.78it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 PEG val sites:  83%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u258e | 10/12 [00:03<00:00,  4.02it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 PEG val sites:  92%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u258f| 11/12 [00:03<00:00,  3.95it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 PEG val sites: 100%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588| 12/12 [00:04<00:00,  3.13it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "NAT4 PEG val sites: 100%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588| 12/12 [00:04<00:00,  2.98it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PEG: 398 cells, 593 stims\n",
      "  auditory: 339 / 398\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  coverage: 236,014 / 236,014 valid pairs (100.0%)\n",
      "  animals:  {'ARM': 398}\n"
     ]
    }
   ],
   "source": [
    "ds_peg = NAT4Dataset(area='PEG', download=True, subset='all')\n",
    "print(f\"PEG: {ds_peg.N_neurons} cells, {len(ds_peg.stims)} stims\")\n",
    "peg_auditory = sum(1 for n in ds_peg.nrn_meta if n['auditory'])\n",
    "print(f\"  auditory: {peg_auditory} / {ds_peg.N_neurons}\")\n",
    "peg_total = ds_peg.nrn_masks.sum().item()\n",
    "peg_grid = ds_peg.nrn_masks.numel()\n",
    "print(f\"  coverage: {peg_total:,} / {peg_grid:,} valid pairs ({100*peg_total/peg_grid:.1f}%)\")\n",
    "print(f\"  animals:  {dict(Counter(n['animal'] for n in ds_peg.nrn_meta))}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "198e15c9",
   "metadata": {},
   "source": [
    "## 9. Cross-area concatenation\n",
    "\n",
    "`concat_neural_datasets([a1, peg])` builds a block-diagonal joint\n",
    "dataset along both stim and neuron axes \u2014 A1's neurons see only A1's\n",
    "stimuli (and vice versa) via NaN sentinels. A single population model\n",
    "over the union jointly fits both areas. See\n",
    "[`dataset_concatenation.md`](../docs/_source/md/dataset_concatenation.md)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "03c1f310",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-02T16:49:27.970792Z",
     "iopub.status.busy": "2026-05-02T16:49:27.970660Z",
     "iopub.status.idle": "2026-05-02T16:49:49.574871Z",
     "shell.execute_reply": "2026-05-02T16:49:49.574141Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "joint dataset:\n",
      "  cells:  1247\n",
      "  stims:  1186\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  coverage: 738,877 / 1,478,942 valid pairs (50.0%)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  cross-area sanity: 738,877 == 738,877\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "joint = concat_neural_datasets([ds, ds_peg])\n",
    "print(f\"joint dataset:\")\n",
    "print(f\"  cells:  {joint.N_neurons}\")\n",
    "print(f\"  stims:  {len(joint.stims)}\")\n",
    "joint_total = joint.nrn_masks.sum().item()\n",
    "joint_grid = joint.nrn_masks.numel()\n",
    "print(f\"  coverage: {joint_total:,} / {joint_grid:,} valid pairs ({100*joint_total/joint_grid:.1f}%)\")\n",
    "print(f\"  cross-area sanity: {ds.nrn_masks.sum().item() + ds_peg.nrn_masks.sum().item():,} == {joint_total:,}\")\n",
    "\n",
    "# verify the block-diagonal structure visually for a small subsample\n",
    "fig, ax = plt.subplots(figsize=(9, 4))\n",
    "sub = joint.nrn_masks[::6, ::8]    # every 6th stim, every 8th cell\n",
    "ax.imshow(sub.numpy().T, aspect='auto', cmap='Greys', interpolation='nearest')\n",
    "ax.axvline(len(ds.stims) // 6 - 0.5, color='red', lw=1.5, label='A1 | PEG stim boundary')\n",
    "ax.axhline(ds.N_neurons // 8 - 0.5, color='blue', lw=1.5, label='A1 | PEG cell boundary')\n",
    "ax.set_xlabel(\"stim (subsampled)\")\n",
    "ax.set_ylabel(\"cell (subsampled)\")\n",
    "ax.set_title(\"Block-diagonal A1 \u2295 PEG joint mask\")\n",
    "ax.legend(loc='lower right')\n",
    "plt.tight_layout(); plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "888b6a93",
   "metadata": {},
   "source": [
    "## What's next\n",
    "\n",
    "- **End-to-end fit on NAT4** \u2014 the same `Fitter`/metrics stack from\n",
    "  [`fit_ns1_statenet.ipynb`](fit_ns1_statenet.ipynb) ports to NAT4\n",
    "  unchanged. The non-trivial bit is the est/val split: NAT4 already\n",
    "  ships it, so use `subset='est'` for the train/val loaders and\n",
    "  `subset='val'` for held-out evaluation.\n",
    "- **Cross-species pooling** \u2014 replace the A1+PEG concat above with\n",
    "  e.g. NS1+NAT4_A1 to fit ferret + ferret jointly, or with CRCNS_AA1\n",
    "  for cross-species.\n",
    "- **Stim selection tricks** \u2014 the bidirectional rule (\u00a75) makes\n",
    "  per-stim-class evaluation a one-liner. Useful when reporting cc_norm\n",
    "  on a particular stimulus class.\n"
   ]
  }
 ],
 "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": "3.11.15"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}