Rich's_mice_dataset_of_2017_08_19.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/mash3133/PharmaHack2024/blob/main/Rich's_mice_dataset_of_2017_08_19.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "04a31379"
      },
      "source": [
        "# Reading NWB files from Dandiset 000579: Two-photon calcium imaging of mouse posterior cortical areas during dynamic navigation decisions\n",
        "\n",
        "---\n",
        "This is the example notebook for reading the NWB files from the [Dandiset 000579: \"Two-photon calcium imaging of mouse posterior cortical areas during dynamic navigation decisions.](https://dandiarchive.org/dandiset/000579/draft)\"\n",
        "\n",
        "*Created by [Shih-Yi Tseng](https://github.com/sytseng) from the [Harvey Lab](https://harveylab.hms.harvard.edu/) at Harvard Medical School*\n",
        "\n",
        "*Last updated: 7/21/2023*\n"
      ],
      "id": "04a31379"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "636280b9"
      },
      "source": [
        "#### Related publication\n",
        "Tseng, S.Y., Chettih, S.N., Arlt, C., Barroso-Luque, R. and Harvey, C.D., 2022. Shared and specialized coding across posterior cortical areas for dynamic navigation decisions. Neuron, 110(15), pp.2484-2502. [DOI:10.1016/j.neuron.2022.05.012](https://www.sciencedirect.com/science/article/pii/S0896627322004536?via%3Dihub).\n",
        "\n",
        "#### Dandiset description\n",
        "\n",
        "The dataset contains calcium activity of >200,000 neurons recorded from 6 different cortical areas in mouse posterior cortex L2/3 and L5 using two-photon imaging, including V1 and secondary visual areas (AM and PM), retrosplenial cortex (RSC) and posterior parietal cortex (visA and MM), while the mice were performing a flexible decision-making task based on rule-switching during virtual navigation. There are total 300 behavior + imaging sessions collected from 8 mice. The neurons in each experiment have been registered into the Allen Institute Mouse Common Coordinate Framework (CCF) based on widefield retinotopy. In addition, these neurons contain fluorescent labels of retroAAV injected in one of the two sets of projection targets: an anterior part of anterior cingulate cortex/secondary motor cortex (ACC/M2) and striatum, or a posterior part of ACC/M2 and orbitofrontal cortex (OFC).\n",
        "\n",
        "#### Task description\n",
        "The mouse ran down a virtual Y-maze and used visual cues on the wall to guide its choice into one of the two maze arms. After the reaching the end of the arm, the mouse was presented with a feedback delay, followed by a visual feedback indicating the correctness of its choice, and then received reward for correct trials or entered inter-trial intervals for incorrect trials (shown in **B**). The rewarded associations between cue and choice are determined by two rules (rule A or rule B as indicated in **A**), and the rule swicthed in blocks of 100-175 trials without explicit signaling multiple times within a single session. The mouse's performance usually dropped below chance level (fraction correct < 0.5) and recovered to expert level after 30-50 trials after rule switches (shown in **C**)."
      ],
      "id": "636280b9"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5028b5c9"
      },
      "source": [
        "![Schematic.png](data:image/png;base64,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)"
      ],
      "id": "5028b5c9"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "d0cc7b7f"
      },
      "source": [
        "The two following sections describe how to read, extract and visualize information in the two types of NWB files in the Dandiset: *single session NWB files with behavior + ophys* in **Section A** and mouse-specific *NWB files with widefield retinotopy and two-photon overview images* in **Section B**."
      ],
      "id": "d0cc7b7f"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "c50c4124"
      },
      "source": [
        "# Section A: Read single session NWB file (behavior+ophys)\n",
        "\n",
        "This section describes how to read and extract information contained in the single session NWB files with behavior + ophys, as well as some basic visualization of the data.\n",
        "\n",
        "## Install PyNWB and dandi\n",
        "\n",
        "First of all, install [PyNWB](https://pynwb.readthedocs.io/en/stable/index.html), the Python package for the Neurodata Without Borders (NWB) data format in your local environment.\n",
        "\n",
        "`pip install -U pynwb` or `conda install -c conda-forge pynwb`\n",
        "\n",
        "You can also install the DANDI CLI into your Python environment, which allows you to interact with the DANDI Archive database with their CLI tools. See the [DANDI handbook](https://www.dandiarchive.org/handbook/) for more instructions.\n",
        "\n",
        "`pip install -U dandi`\n",
        "\n",
        "Alternatively, you can use the web-based data portal to download files from Dandisets.\n",
        "\n",
        "## Install NWB extension for custom lab meta data\n",
        "\n",
        "Next, install the NWB extnesion for custom lab meta data from here: https://github.com/sytseng/ndx-harvey-swac and install it in your environment, or call\n",
        "\n",
        "`pip install git+https://github.com/sytseng/ndx-harvey-swac.git`"
      ],
      "id": "c50c4124"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "background_save": true,
          "base_uri": "https://localhost:8080/"
        },
        "id": "htmzCpKi8GeL",
        "outputId": "7222540b-02dc-4ac9-9361-d840eee99ad3"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: dandi in /usr/local/lib/python3.10/dist-packages (0.60.0)\n",
            "Requirement already satisfied: bidsschematools~=0.7.0 in /usr/local/lib/python3.10/dist-packages (from dandi) (0.7.2)\n",
            "Requirement already satisfied: click>=7.1 in /usr/local/lib/python3.10/dist-packages (from dandi) (8.1.7)\n",
            "Requirement already satisfied: click-didyoumean in /usr/local/lib/python3.10/dist-packages (from dandi) (0.3.0)\n",
            "Requirement already satisfied: dandischema<0.11,>=0.9.0 in /usr/local/lib/python3.10/dist-packages (from dandi) (0.10.0)\n",
            "Requirement already satisfied: etelemetry>=0.2.2 in /usr/local/lib/python3.10/dist-packages (from dandi) (0.3.1)\n",
            "Requirement already satisfied: fasteners in /usr/local/lib/python3.10/dist-packages (from dandi) (0.19)\n",
            "Requirement already satisfied: fscacher>=0.3.0 in /usr/local/lib/python3.10/dist-packages (from dandi) (0.4.0)\n",
            "Requirement already satisfied: hdmf!=3.5.0 in /usr/local/lib/python3.10/dist-packages (from dandi) (3.12.2)\n",
            "Requirement already satisfied: humanize in /usr/local/lib/python3.10/dist-packages (from dandi) (4.7.0)\n",
            "Requirement already satisfied: interleave~=0.1 in /usr/local/lib/python3.10/dist-packages (from dandi) (0.2.1)\n",
            "Requirement already satisfied: joblib in /usr/local/lib/python3.10/dist-packages (from dandi) (1.3.2)\n",
            "Requirement already satisfied: keyring!=23.9.0 in /usr/lib/python3/dist-packages (from dandi) (23.5.0)\n",
            "Requirement already satisfied: keyrings.alt in /usr/local/lib/python3.10/dist-packages (from dandi) (5.0.0)\n",
            "Requirement already satisfied: packaging in /usr/local/lib/python3.10/dist-packages (from dandi) (24.0)\n",
            "Requirement already satisfied: platformdirs in /usr/local/lib/python3.10/dist-packages (from dandi) (4.2.0)\n",
            "Requirement already satisfied: pycryptodomex in /usr/local/lib/python3.10/dist-packages (from dandi) (3.20.0)\n",
            "Requirement already satisfied: pydantic~=2.0 in /usr/local/lib/python3.10/dist-packages (from dandi) (2.6.4)\n",
            "Requirement already satisfied: pynwb!=1.1.0,!=2.3.0,>=1.0.3 in /usr/local/lib/python3.10/dist-packages (from dandi) (2.6.0)\n",
            "Requirement already satisfied: nwbinspector!=0.4.32,>=0.4.28 in /usr/local/lib/python3.10/dist-packages (from dandi) (0.4.33)\n",
            "Requirement already satisfied: pyout!=0.6.0,>=0.5 in /usr/local/lib/python3.10/dist-packages (from dandi) (0.7.3)\n",
            "Requirement already satisfied: python-dateutil in /usr/local/lib/python3.10/dist-packages (from dandi) (2.8.2)\n",
            "Requirement already satisfied: requests~=2.20 in /usr/local/lib/python3.10/dist-packages (from dandi) (2.31.0)\n",
            "Requirement already satisfied: ruamel.yaml<1,>=0.15 in /usr/local/lib/python3.10/dist-packages (from dandi) (0.18.6)\n",
            "Requirement already satisfied: semantic-version in /usr/local/lib/python3.10/dist-packages (from dandi) (2.10.0)\n",
            "Requirement already satisfied: tenacity in /usr/local/lib/python3.10/dist-packages (from dandi) (8.2.3)\n",
            "Requirement already satisfied: zarr~=2.10 in /usr/local/lib/python3.10/dist-packages (from dandi) (2.17.1)\n",
            "Requirement already satisfied: zarr-checksum~=0.4.0 in /usr/local/lib/python3.10/dist-packages (from dandi) (0.4.0)\n",
            "Requirement already satisfied: pyyaml in /usr/local/lib/python3.10/dist-packages (from bidsschematools~=0.7.0->dandi) (6.0.1)\n",
            "Requirement already satisfied: jsonschema[format] in /usr/local/lib/python3.10/dist-packages (from dandischema<0.11,>=0.9.0->dandi) (4.19.2)\n",
            "Requirement already satisfied: ci-info>=0.2 in /usr/local/lib/python3.10/dist-packages (from etelemetry>=0.2.2->dandi) (0.3.0)\n",
            "Requirement already satisfied: appdirs==1.* in /usr/local/lib/python3.10/dist-packages (from fscacher>=0.3.0->dandi) (1.4.4)\n",
            "Requirement already satisfied: h5py>=2.10 in /usr/local/lib/python3.10/dist-packages (from hdmf!=3.5.0->dandi) (3.9.0)\n",
            "Requirement already satisfied: numpy>=1.18 in /usr/local/lib/python3.10/dist-packages (from hdmf!=3.5.0->dandi) (1.25.2)\n",
            "Requirement already satisfied: pandas>=1.0.5 in /usr/local/lib/python3.10/dist-packages (from hdmf!=3.5.0->dandi) (1.5.3)\n",
            "Requirement already satisfied: scipy>=1.4 in /usr/local/lib/python3.10/dist-packages (from hdmf!=3.5.0->dandi) (1.11.4)\n",
            "Requirement already satisfied: fsspec in /usr/local/lib/python3.10/dist-packages (from nwbinspector!=0.4.32,>=0.4.28->dandi) (2024.3.0)\n",
            "Requirement already satisfied: s3fs in /usr/local/lib/python3.10/dist-packages (from nwbinspector!=0.4.32,>=0.4.28->dandi) (2024.3.0)\n",
            "Requirement already satisfied: aiohttp in /usr/local/lib/python3.10/dist-packages (from nwbinspector!=0.4.32,>=0.4.28->dandi) (3.9.3)\n",
            "Requirement already satisfied: natsort in /usr/local/lib/python3.10/dist-packages (from nwbinspector!=0.4.32,>=0.4.28->dandi) (8.4.0)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.10/dist-packages (from nwbinspector!=0.4.32,>=0.4.28->dandi) (4.66.2)\n",
            "Requirement already satisfied: isodate in /usr/local/lib/python3.10/dist-packages (from nwbinspector!=0.4.32,>=0.4.28->dandi) (0.6.1)\n",
            "Requirement already satisfied: annotated-types>=0.4.0 in /usr/local/lib/python3.10/dist-packages (from pydantic~=2.0->dandi) (0.6.0)\n",
            "Requirement already satisfied: pydantic-core==2.16.3 in /usr/local/lib/python3.10/dist-packages (from pydantic~=2.0->dandi) (2.16.3)\n",
            "Requirement already satisfied: typing-extensions>=4.6.1 in /usr/local/lib/python3.10/dist-packages (from pydantic~=2.0->dandi) (4.10.0)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.10/dist-packages (from pynwb!=1.1.0,!=2.3.0,>=1.0.3->dandi) (67.7.2)\n",
            "Requirement already satisfied: blessed in /usr/local/lib/python3.10/dist-packages (from pyout!=0.6.0,>=0.5->dandi) (1.20.0)\n",
            "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil->dandi) (1.16.0)\n",
            "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests~=2.20->dandi) (3.3.2)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests~=2.20->dandi) (3.6)\n",
            "Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests~=2.20->dandi) (2.0.7)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests~=2.20->dandi) (2024.2.2)\n",
            "Requirement already satisfied: ruamel.yaml.clib>=0.2.7 in /usr/local/lib/python3.10/dist-packages (from ruamel.yaml<1,>=0.15->dandi) (0.2.8)\n",
            "Requirement already satisfied: asciitree in /usr/local/lib/python3.10/dist-packages (from zarr~=2.10->dandi) (0.3.3)\n",
            "Requirement already satisfied: numcodecs>=0.10.0 in /usr/local/lib/python3.10/dist-packages (from zarr~=2.10->dandi) (0.12.1)\n",
            "Requirement already satisfied: jaraco.classes in /usr/local/lib/python3.10/dist-packages (from keyrings.alt->dandi) (3.3.1)\n",
            "Requirement already satisfied: attrs>=22.2.0 in /usr/local/lib/python3.10/dist-packages (from jsonschema[format]->dandischema<0.11,>=0.9.0->dandi) (23.2.0)\n",
            "Requirement already satisfied: jsonschema-specifications>=2023.03.6 in /usr/local/lib/python3.10/dist-packages (from jsonschema[format]->dandischema<0.11,>=0.9.0->dandi) (2023.12.1)\n",
            "Requirement already satisfied: referencing>=0.28.4 in /usr/local/lib/python3.10/dist-packages (from jsonschema[format]->dandischema<0.11,>=0.9.0->dandi) (0.33.0)\n",
            "Requirement already satisfied: rpds-py>=0.7.1 in /usr/local/lib/python3.10/dist-packages (from jsonschema[format]->dandischema<0.11,>=0.9.0->dandi) (0.18.0)\n",
            "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas>=1.0.5->hdmf!=3.5.0->dandi) (2023.4)\n",
            "Requirement already satisfied: email-validator>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from pydantic~=2.0->dandi) (2.1.1)\n",
            "Requirement already satisfied: aiosignal>=1.1.2 in /usr/local/lib/python3.10/dist-packages (from aiohttp->nwbinspector!=0.4.32,>=0.4.28->dandi) (1.3.1)\n",
            "Requirement already satisfied: frozenlist>=1.1.1 in /usr/local/lib/python3.10/dist-packages (from aiohttp->nwbinspector!=0.4.32,>=0.4.28->dandi) (1.4.1)\n",
            "Requirement already satisfied: multidict<7.0,>=4.5 in /usr/local/lib/python3.10/dist-packages (from aiohttp->nwbinspector!=0.4.32,>=0.4.28->dandi) (6.0.5)\n",
            "Requirement already satisfied: yarl<2.0,>=1.0 in /usr/local/lib/python3.10/dist-packages (from aiohttp->nwbinspector!=0.4.32,>=0.4.28->dandi) (1.9.4)\n",
            "Requirement already satisfied: async-timeout<5.0,>=4.0 in /usr/local/lib/python3.10/dist-packages (from aiohttp->nwbinspector!=0.4.32,>=0.4.28->dandi) (4.0.3)\n",
            "Requirement already satisfied: wcwidth>=0.1.4 in /usr/local/lib/python3.10/dist-packages (from blessed->pyout!=0.6.0,>=0.5->dandi) (0.2.13)\n",
            "Requirement already satisfied: more-itertools in /usr/local/lib/python3.10/dist-packages (from jaraco.classes->keyrings.alt->dandi) (10.1.0)\n",
            "Requirement already satisfied: fqdn in /usr/local/lib/python3.10/dist-packages (from jsonschema[format]->dandischema<0.11,>=0.9.0->dandi) (1.5.1)\n",
            "Requirement already satisfied: isoduration in /usr/local/lib/python3.10/dist-packages (from jsonschema[format]->dandischema<0.11,>=0.9.0->dandi) (20.11.0)\n",
            "Requirement already satisfied: jsonpointer>1.13 in /usr/local/lib/python3.10/dist-packages (from jsonschema[format]->dandischema<0.11,>=0.9.0->dandi) (2.4)\n",
            "Requirement already satisfied: rfc3339-validator in /usr/local/lib/python3.10/dist-packages (from jsonschema[format]->dandischema<0.11,>=0.9.0->dandi) (0.1.4)\n",
            "Requirement already satisfied: rfc3987 in /usr/local/lib/python3.10/dist-packages (from jsonschema[format]->dandischema<0.11,>=0.9.0->dandi) (1.3.8)\n",
            "Requirement already satisfied: uri-template in /usr/local/lib/python3.10/dist-packages (from jsonschema[format]->dandischema<0.11,>=0.9.0->dandi) (1.3.0)\n",
            "Requirement already satisfied: webcolors>=1.11 in /usr/local/lib/python3.10/dist-packages (from jsonschema[format]->dandischema<0.11,>=0.9.0->dandi) (1.13)\n",
            "Requirement already satisfied: aiobotocore<3.0.0,>=2.5.4 in /usr/local/lib/python3.10/dist-packages (from s3fs->nwbinspector!=0.4.32,>=0.4.28->dandi) (2.12.1)\n",
            "Requirement already satisfied: botocore<1.34.52,>=1.34.41 in /usr/local/lib/python3.10/dist-packages (from aiobotocore<3.0.0,>=2.5.4->s3fs->nwbinspector!=0.4.32,>=0.4.28->dandi) (1.34.51)\n",
            "Requirement already satisfied: wrapt<2.0.0,>=1.10.10 in /usr/local/lib/python3.10/dist-packages (from aiobotocore<3.0.0,>=2.5.4->s3fs->nwbinspector!=0.4.32,>=0.4.28->dandi) (1.14.1)\n",
            "Requirement already satisfied: aioitertools<1.0.0,>=0.5.1 in /usr/local/lib/python3.10/dist-packages (from aiobotocore<3.0.0,>=2.5.4->s3fs->nwbinspector!=0.4.32,>=0.4.28->dandi) (0.11.0)\n",
            "Requirement already satisfied: dnspython>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from email-validator>=2.0.0->pydantic~=2.0->dandi) (2.6.1)\n",
            "Requirement already satisfied: arrow>=0.15.0 in /usr/local/lib/python3.10/dist-packages (from isoduration->jsonschema[format]->dandischema<0.11,>=0.9.0->dandi) (1.3.0)\n",
            "Requirement already satisfied: types-python-dateutil>=2.8.10 in /usr/local/lib/python3.10/dist-packages (from arrow>=0.15.0->isoduration->jsonschema[format]->dandischema<0.11,>=0.9.0->dandi) (2.9.0.20240316)\n",
            "Requirement already satisfied: jmespath<2.0.0,>=0.7.1 in /usr/local/lib/python3.10/dist-packages (from botocore<1.34.52,>=1.34.41->aiobotocore<3.0.0,>=2.5.4->s3fs->nwbinspector!=0.4.32,>=0.4.28->dandi) (1.0.1)\n",
            "Requirement already satisfied: pynwb in /usr/local/lib/python3.10/dist-packages (2.6.0)\n",
            "Requirement already satisfied: h5py>=2.10 in /usr/local/lib/python3.10/dist-packages (from pynwb) (3.9.0)\n",
            "Requirement already satisfied: hdmf>=3.12.2 in /usr/local/lib/python3.10/dist-packages (from pynwb) (3.12.2)\n",
            "Requirement already satisfied: numpy>=1.16 in /usr/local/lib/python3.10/dist-packages (from pynwb) (1.25.2)\n",
            "Requirement already satisfied: pandas>=1.1.5 in /usr/local/lib/python3.10/dist-packages (from pynwb) (1.5.3)\n",
            "Requirement already satisfied: python-dateutil>=2.7.3 in /usr/local/lib/python3.10/dist-packages (from pynwb) (2.8.2)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.10/dist-packages (from pynwb) (67.7.2)\n",
            "Requirement already satisfied: jsonschema>=2.6.0 in /usr/local/lib/python3.10/dist-packages (from hdmf>=3.12.2->pynwb) (4.19.2)\n",
            "Requirement already satisfied: ruamel-yaml>=0.16 in /usr/local/lib/python3.10/dist-packages (from hdmf>=3.12.2->pynwb) (0.18.6)\n",
            "Requirement already satisfied: scipy>=1.4 in /usr/local/lib/python3.10/dist-packages (from hdmf>=3.12.2->pynwb) (1.11.4)\n",
            "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas>=1.1.5->pynwb) (2023.4)\n",
            "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.7.3->pynwb) (1.16.0)\n",
            "Requirement already satisfied: attrs>=22.2.0 in /usr/local/lib/python3.10/dist-packages (from jsonschema>=2.6.0->hdmf>=3.12.2->pynwb) (23.2.0)\n",
            "Requirement already satisfied: jsonschema-specifications>=2023.03.6 in /usr/local/lib/python3.10/dist-packages (from jsonschema>=2.6.0->hdmf>=3.12.2->pynwb) (2023.12.1)\n",
            "Requirement already satisfied: referencing>=0.28.4 in /usr/local/lib/python3.10/dist-packages (from jsonschema>=2.6.0->hdmf>=3.12.2->pynwb) (0.33.0)\n",
            "Requirement already satisfied: rpds-py>=0.7.1 in /usr/local/lib/python3.10/dist-packages (from jsonschema>=2.6.0->hdmf>=3.12.2->pynwb) (0.18.0)\n",
            "Requirement already satisfied: ruamel.yaml.clib>=0.2.7 in /usr/local/lib/python3.10/dist-packages (from ruamel-yaml>=0.16->hdmf>=3.12.2->pynwb) (0.2.8)\n",
            "Collecting git+https://github.com/sytseng/ndx-harvey-swac.git\n",
            "  Cloning https://github.com/sytseng/ndx-harvey-swac.git to /tmp/pip-req-build-sjv2vqn5\n",
            "  Running command git clone --filter=blob:none --quiet https://github.com/sytseng/ndx-harvey-swac.git /tmp/pip-req-build-sjv2vqn5\n",
            "  Resolved https://github.com/sytseng/ndx-harvey-swac.git to commit c82b84cbf7c9d3e8c320c439dab3bf2fd48f0ae8\n",
            "  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Requirement already satisfied: pynwb<3,>=1.5.0 in /usr/local/lib/python3.10/dist-packages (from ndx-harvey-swac==0.1.0) (2.6.0)\n",
            "Requirement already satisfied: hdmf<4,>=2.5.6 in /usr/local/lib/python3.10/dist-packages (from ndx-harvey-swac==0.1.0) (3.12.2)\n",
            "Requirement already satisfied: h5py>=2.10 in /usr/local/lib/python3.10/dist-packages (from hdmf<4,>=2.5.6->ndx-harvey-swac==0.1.0) (3.9.0)\n",
            "Requirement already satisfied: jsonschema>=2.6.0 in /usr/local/lib/python3.10/dist-packages (from hdmf<4,>=2.5.6->ndx-harvey-swac==0.1.0) (4.19.2)\n",
            "Requirement already satisfied: numpy>=1.18 in /usr/local/lib/python3.10/dist-packages (from hdmf<4,>=2.5.6->ndx-harvey-swac==0.1.0) (1.25.2)\n",
            "Requirement already satisfied: pandas>=1.0.5 in /usr/local/lib/python3.10/dist-packages (from hdmf<4,>=2.5.6->ndx-harvey-swac==0.1.0) (1.5.3)\n",
            "Requirement already satisfied: ruamel-yaml>=0.16 in /usr/local/lib/python3.10/dist-packages (from hdmf<4,>=2.5.6->ndx-harvey-swac==0.1.0) (0.18.6)\n",
            "Requirement already satisfied: scipy>=1.4 in /usr/local/lib/python3.10/dist-packages (from hdmf<4,>=2.5.6->ndx-harvey-swac==0.1.0) (1.11.4)\n",
            "Requirement already satisfied: python-dateutil>=2.7.3 in /usr/local/lib/python3.10/dist-packages (from pynwb<3,>=1.5.0->ndx-harvey-swac==0.1.0) (2.8.2)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.10/dist-packages (from pynwb<3,>=1.5.0->ndx-harvey-swac==0.1.0) (67.7.2)\n",
            "Requirement already satisfied: attrs>=22.2.0 in /usr/local/lib/python3.10/dist-packages (from jsonschema>=2.6.0->hdmf<4,>=2.5.6->ndx-harvey-swac==0.1.0) (23.2.0)\n",
            "Requirement already satisfied: jsonschema-specifications>=2023.03.6 in /usr/local/lib/python3.10/dist-packages (from jsonschema>=2.6.0->hdmf<4,>=2.5.6->ndx-harvey-swac==0.1.0) (2023.12.1)\n",
            "Requirement already satisfied: referencing>=0.28.4 in /usr/local/lib/python3.10/dist-packages (from jsonschema>=2.6.0->hdmf<4,>=2.5.6->ndx-harvey-swac==0.1.0) (0.33.0)\n",
            "Requirement already satisfied: rpds-py>=0.7.1 in /usr/local/lib/python3.10/dist-packages (from jsonschema>=2.6.0->hdmf<4,>=2.5.6->ndx-harvey-swac==0.1.0) (0.18.0)\n",
            "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas>=1.0.5->hdmf<4,>=2.5.6->ndx-harvey-swac==0.1.0) (2023.4)\n",
            "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.7.3->pynwb<3,>=1.5.0->ndx-harvey-swac==0.1.0) (1.16.0)\n",
            "Requirement already satisfied: ruamel.yaml.clib>=0.2.7 in /usr/local/lib/python3.10/dist-packages (from ruamel-yaml>=0.16->hdmf<4,>=2.5.6->ndx-harvey-swac==0.1.0) (0.2.8)\n"
          ]
        }
      ],
      "source": [
        "!pip install -U dandi\n",
        "!pip install -U pynwb\n",
        "!pip install git+https://github.com/sytseng/ndx-harvey-swac.git"
      ],
      "id": "htmzCpKi8GeL"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bbaa5094"
      },
      "source": [
        "Then load libaries:"
      ],
      "id": "bbaa5094"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "background_save": true
        },
        "id": "95905316"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "\n",
        "from pynwb import NWBHDF5IO\n",
        "from ndx_harvey_swac import LabMetaDataSession"
      ],
      "id": "95905316"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3bdce6cb"
      },
      "source": [
        "## Download an example NWB file\n",
        "\n",
        "Download the NWB file from DANDI dataset (https://dandiarchive.org/dandiset/000579/draft). For example, download the session file for mouse 9 session date 2017/08/19 (Go to FILES on the right panel -> click on the folder for sub-9 -> find the file named *sub-9_ses-mouse-9-session-date-2017-08-19-area-visA-L23-multi-plane-imaging_behavior+ophys.nwb*). You can download it directly from the DANDI website, or use DANDI CLI in the command line window:\n",
        "\n",
        "`dandi download https://api.dandiarchive.org/api/assets/0dbf026f-1c13-45ed-874f-83c07a742d32/download/`"
      ],
      "id": "3bdce6cb"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "background_save": true,
          "base_uri": "https://localhost:8080/"
        },
        "id": "1CxwAR068lYG",
        "outputId": "a49c4842-a69f-4df2-de8e-aa2144348efb"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "PATH                                                                                          SIZE      DONE    DONE% CHECKSUM STATUS  MESSAGE          \n",
            "sub-9_ses-mouse-9-session-date-2017-08-17-area-RSC-L23-multi-plane-imaging_behavior+ophys.nwb                                  error   FileExistsError  \n",
            "Summary:                                                                                      0 Bytes   0 Bytes                1 error 1 FileExistsError\n",
            "                                                                                              +611.9 MB 0.00%                                           \n",
            "2024-03-17 16:44:16,235 [    INFO] Logs saved in /root/.local/state/dandi-cli/log/20240317164415Z-53613.log\n",
            "PATH                                                                                          SIZE      DONE    DONE% CHECKSUM STATUS  MESSAGE          \n",
            "sub-9_ses-mouse-9-session-date-2017-08-07-area-RSC-L23-multi-plane-imaging_behavior+ophys.nwb                                  error   FileExistsError  \n",
            "Summary:                                                                                      0 Bytes   0 Bytes                1 error 1 FileExistsError\n",
            "                                                                                              +657.2 MB 0.00%                                           \n",
            "2024-03-17 16:44:18,424 [    INFO] Logs saved in /root/.local/state/dandi-cli/log/20240317164417Z-53625.log\n",
            "PATH                                                                                         SIZE      DONE    DONE% CHECKSUM STATUS  MESSAGE          \n",
            "sub-9_ses-mouse-9-session-date-2017-08-09-area-MM-L23-multi-plane-imaging_behavior+ophys.nwb                                  error   FileExistsError  \n",
            "Summary:                                                                                     0 Bytes   0 Bytes                1 error 1 FileExistsError\n",
            "                                                                                             +504.0 MB 0.00%                                           \n",
            "2024-03-17 16:44:20,568 [    INFO] Logs saved in /root/.local/state/dandi-cli/log/20240317164419Z-53637.log\n",
            "PATH                                                                                           SIZE      DONE    DONE% CHECKSUM STATUS  MESSAGE          \n",
            "sub-9_ses-mouse-9-session-date-2017-08-11-area-visA-L23-multi-plane-imaging_behavior+ophys.nwb                                  error   FileExistsError  \n",
            "Summary:                                                                                       0 Bytes   0 Bytes                1 error 1 FileExistsError\n",
            "                                                                                               +556.4 MB 0.00%                                           \n",
            "2024-03-17 16:44:22,847 [    INFO] Logs saved in /root/.local/state/dandi-cli/log/20240317164421Z-53653.log\n"
          ]
        }
      ],
      "source": [
        "!dandi download https://api.dandiarchive.org/api/assets/0dbf026f-1c13-45ed-874f-83c07a742d32/download/\n",
        "!dandi download https://api.dandiarchive.org/api/assets/240d4fa5-278a-422f-aaa0-5afff381bb57/download/\n",
        "!dandi download https://api.dandiarchive.org/api/assets/4ddd7455-ab78-402c-92c8-a55ce6eb689b/download/\n",
        "!dandi download https://api.dandiarchive.org/api/assets/3362af75-91fc-4e40-8536-aabcb74f8f15/download/"
      ],
      "id": "1CxwAR068lYG"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "background_save": true
        },
        "id": "11cd7615"
      },
      "outputs": [],
      "source": [
        "# specify the file path in your local file system (PLEASE CHANGE THIS!)\n",
        "filepaths = [\n",
        "    \"sub-9_ses-mouse-9-session-date-2017-08-17-area-RSC-L23-multi-plane-imaging_behavior+ophys.nwb\",\n",
        "    \"sub-9_ses-mouse-9-session-date-2017-08-07-area-RSC-L23-multi-plane-imaging_behavior+ophys.nwb\",\n",
        "    \"sub-9_ses-mouse-9-session-date-2017-08-09-area-MM-L23-multi-plane-imaging_behavior+ophys.nwb\",\n",
        "    \"sub-9_ses-mouse-9-session-date-2017-08-11-area-visA-L23-multi-plane-imaging_behavior+ophys.nwb\",\n",
        "]"
      ],
      "id": "11cd7615"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 304
        },
        "id": "_GcNRgHHBNFS",
        "outputId": "da82a007-0816-401b-960c-9d53590767d6"
      },
      "outputs": [
        {
          "ename": "KeyboardInterrupt",
          "evalue": "",
          "output_type": "error",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-161-d5df0069828e>\u001b[0m in \u001b[0;36m<cell line: 2>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mgoogle\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcolab\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mdrive\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mdrive\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmount\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'/content/drive'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/google/colab/drive.py\u001b[0m in \u001b[0;36mmount\u001b[0;34m(mountpoint, force_remount, timeout_ms, readonly)\u001b[0m\n\u001b[1;32m     98\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mmount\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmountpoint\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mforce_remount\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtimeout_ms\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m120000\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreadonly\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     99\u001b[0m   \u001b[0;34m\"\"\"Mount your Google Drive at the specified mountpoint path.\"\"\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 100\u001b[0;31m   return _mount(\n\u001b[0m\u001b[1;32m    101\u001b[0m       \u001b[0mmountpoint\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    102\u001b[0m       \u001b[0mforce_remount\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mforce_remount\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/google/colab/drive.py\u001b[0m in \u001b[0;36m_mount\u001b[0;34m(mountpoint, force_remount, timeout_ms, ephemeral, readonly)\u001b[0m\n\u001b[1;32m    131\u001b[0m   )\n\u001b[1;32m    132\u001b[0m   \u001b[0;32mif\u001b[0m \u001b[0mephemeral\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 133\u001b[0;31m     _message.blocking_request(\n\u001b[0m\u001b[1;32m    134\u001b[0m         \u001b[0;34m'request_auth'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrequest\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m{\u001b[0m\u001b[0;34m'authType'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;34m'dfs_ephemeral'\u001b[0m\u001b[0;34m}\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtimeout_sec\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    135\u001b[0m     )\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/google/colab/_message.py\u001b[0m in \u001b[0;36mblocking_request\u001b[0;34m(request_type, request, timeout_sec, parent)\u001b[0m\n\u001b[1;32m    174\u001b[0m       \u001b[0mrequest_type\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrequest\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mparent\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mparent\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mexpect_reply\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    175\u001b[0m   )\n\u001b[0;32m--> 176\u001b[0;31m   \u001b[0;32mreturn\u001b[0m \u001b[0mread_reply_from_input\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrequest_id\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtimeout_sec\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/google/colab/_message.py\u001b[0m in \u001b[0;36mread_reply_from_input\u001b[0;34m(message_id, timeout_sec)\u001b[0m\n\u001b[1;32m     94\u001b[0m     \u001b[0mreply\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_read_next_input_message\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     95\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mreply\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0m_NOT_READY\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mreply\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdict\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 96\u001b[0;31m       \u001b[0mtime\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msleep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0.025\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     97\u001b[0m       \u001b[0;32mcontinue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     98\u001b[0m     if (\n",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "id": "_GcNRgHHBNFS"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8aa13ea0"
      },
      "source": [
        "## Read the NWB file\n",
        "Call the `NWBHDF5IO.read()` method to read the NWB data into an `NWBFile` object. Print the `NWBFile` object to inspect its contents."
      ],
      "id": "8aa13ea0"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "8824100a"
      },
      "outputs": [],
      "source": [
        "# Initialize an empty list to store dataframes\n",
        "dfs = []\n",
        "\n",
        "# Loop over each file\n",
        "for filepath in filepaths:\n",
        "    # Read the NWB file\n",
        "    with NWBHDF5IO(filepath, 'r') as io:\n",
        "        nwb = io.read()\n",
        "        # Convert trials data to a dataframe and append to the list\n",
        "        trial_df = nwb.trials.to_dataframe()\n",
        "        dfs.append(trial_df)"
      ],
      "id": "8824100a"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "60038630"
      },
      "outputs": [],
      "source": [
        "# Concatenate all dataframes into a single dataframe\n",
        "combined_df = pd.concat(dfs, ignore_index=True)\n",
        "\n",
        "# Export the combined dataframe to a CSV file\n",
        "combined_df.to_csv('trial_data.csv', index=False)"
      ],
      "id": "60038630"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "c73de741"
      },
      "source": [
        "# Section B: Read NWB file for widefield retinotopy and 2P window overview\n",
        "\n",
        "In this section, we desribe how to read and extract information from the other type of NWB file. In the folder for every mouse, you can find a file named *sub-#_ses-Mouse-#-widefield-retinotopy-and-window-vessel_ophys.nwb*. This it the file for widefield retinotopy and two-photon overview vessel/AAVretro labeling images of each mouse. These mouse-level data was used for registration of the whole cranial window as well as the FOVs for individual behavior+imaging sessions into the Allen Institute Mouse Common Coordinate Framework (CCF).\n",
        "\n",
        "For example, let's download the file for mouse #9.\n",
        "\n",
        "`dandi download https://api.dandiarchive.org/api/assets/cd183c6f-be6e-4773-9320-23a4baed37e6/download/`"
      ],
      "id": "c73de741"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "6Hu4SwKm_T4i"
      },
      "outputs": [],
      "source": [
        "!dandi download https://api.dandiarchive.org/api/assets/cd183c6f-be6e-4773-9320-23a4baed37e6/download/"
      ],
      "id": "6Hu4SwKm_T4i"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "509f49b6"
      },
      "outputs": [],
      "source": [
        "# specify the file path in your local file system (PLEASE CHANGE THIS!)\n",
        "retinoFilepath = 'sub-9_ses-Mouse-9-widefield-retinotopy-and-window-vessel_ophys.nwb'"
      ],
      "id": "509f49b6"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4b255de0"
      },
      "outputs": [],
      "source": [
        "# load the file and display the content\n",
        "io = NWBHDF5IO(retinoFilepath, 'r')\n",
        "nwb_retino = io.read()\n",
        "nwb_retino"
      ],
      "id": "4b255de0"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6535fe34"
      },
      "source": [
        "The `subject`, `surgery` and `virus` notes are identical to those in the single behavior+imaging session NWB file for this mouse."
      ],
      "id": "6535fe34"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "8a74c137"
      },
      "outputs": [],
      "source": [
        "nwb_retino.subject"
      ],
      "id": "8a74c137"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "3c23e7f2"
      },
      "outputs": [],
      "source": [
        "nwb_retino.surgery"
      ],
      "id": "3c23e7f2"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "59fd30c2"
      },
      "outputs": [],
      "source": [
        "nwb_retino.virus"
      ],
      "id": "59fd30c2"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "60e828b8"
      },
      "source": [
        "## Processing modules\n",
        "\n",
        "There are two processing modules in this file: `ophys` that contains static images of the vessel pattern, mTagBFP2 and mScarlet for the whole craniel window in this animal taken with thw two-photon microscope, and `retinotopy` that contains the widefield imaging results for retinotopy."
      ],
      "id": "60e828b8"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "16eebca7"
      },
      "outputs": [],
      "source": [
        "# view content of processing modules\n",
        "nwb_retino.processing"
      ],
      "id": "16eebca7"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "32b263e5"
      },
      "source": [
        "### Whole window vessel image\n",
        "\n",
        "Aside from individual behavior+imaging sessions, we also took an overview image of the vessel pattern under the two-photon microscope with stiching FOVs, which can then be used for registraion of individual FOVs into window coordinate, as well as aligning to the widefield retinotopy results for registraion into Allen CCF.\n",
        "\n",
        "There are two images for the whole window vessel pattern:\n",
        "- **raw_whole_window_vessel_img_2p**: raw image of vessel pattern at brain surface for the whole window (the raw orientation under the 2P microscope, x axis is ~posterior to anterior, y axis is ~lateral (left) to medial (right))\n",
        "- **processed_whole_window_vessel_img_2p**: processed image of raw stitched vessel pattern (removed the overlaping regions, resized to same number of pixels per mm for both axes, and rotated 90 degree + flipped L-R so x axis became ~lateral to medial, y axis became ~posterior to anterior, same orientation as WF vessel)"
      ],
      "id": "32b263e5"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "807bfa4a"
      },
      "outputs": [],
      "source": [
        "# view content of whole_window_vessel_img_2p\n",
        "nwb_retino.processing['ophys']['whole_window_vessel_img_2p']"
      ],
      "id": "807bfa4a"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "3eaeb5fd"
      },
      "outputs": [],
      "source": [
        "# extract vessel pattern images\n",
        "vessel_raw = nwb_retino.processing['ophys']['whole_window_vessel_img_2p']['raw_whole_window_vessel_img_2p'].data[:]\n",
        "vessel_processed = nwb_retino.processing['ophys']['whole_window_vessel_img_2p']['processed_whole_window_vessel_img_2p'].data[:]"
      ],
      "id": "3eaeb5fd"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "314a2d0b"
      },
      "outputs": [],
      "source": [
        "# visualize vessel images\n",
        "fig, ax = plt.subplots(1,2, figsize=(8,4))\n",
        "ax[0].imshow(vessel_raw, vmax = np.percentile(vessel_raw[:], 99.5), cmap='gray')\n",
        "ax[0].set(title = 'Raw vessel image')\n",
        "ax[1].imshow(vessel_processed, vmax = np.percentile(vessel_processed[:], 99.5), cmap='gray')\n",
        "ax[1].set(title = 'Processed vessel image');"
      ],
      "id": "314a2d0b"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "22dc00bb"
      },
      "source": [
        "### Whole window AAVretro label images\n",
        "\n",
        "We also took the overview images of the AAVretro labeling for mTagBFP2 and mScarlet for each cortical layer (L2, L3, L5) under the two-photon microscope with stiching FOVs. Same as the vessel image, there are both the raw images and processed images.\n",
        "\n",
        "**Note that the mTagBFP2 labeling images in L5 were mostly beedthrough signals from the GCaMP; we barely saw any real blue fluorecent signals in L5. Therefore, L5 neurons were always considered not labeled and we are not showing those images below.**"
      ],
      "id": "22dc00bb"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "5510becb"
      },
      "outputs": [],
      "source": [
        "# view content of whole_window_mTagBFP2_img_2p\n",
        "nwb_retino.processing['ophys']['whole_window_mTagBFP2_img_2p']"
      ],
      "id": "5510becb"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "65c725fc"
      },
      "outputs": [],
      "source": [
        "# extract images for mTagBFP2 labeling\n",
        "mTagBFP2_L2_raw = nwb_retino.processing['ophys']['whole_window_mTagBFP2_img_2p']['raw_whole_window_mTagBFP2_img_L2'].data[:]\n",
        "print(mTagBFP2_L2_raw)\n",
        "mTagBFP2_L3_raw = nwb_retino.processing['ophys']['whole_window_mTagBFP2_img_2p']['raw_whole_window_mTagBFP2_img_L3'].data[:]\n",
        "mTagBFP2_L2_processed = nwb_retino.processing['ophys']['whole_window_mTagBFP2_img_2p']['processed_whole_window_mTagBFP2_img_L2'].data[:]\n",
        "print(mTagBFP2_L2_processed)\n",
        "mTagBFP2_L3_processed = nwb_retino.processing['ophys']['whole_window_mTagBFP2_img_2p']['processed_whole_window_mTagBFP2_img_L3'].data[:]\n"
      ],
      "id": "65c725fc"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "c39f577e"
      },
      "outputs": [],
      "source": [
        "# visualize mTagBFP2 labeling images\n",
        "fig, ax = plt.subplots(2,2, figsize=(8,8))\n",
        "ax[0,0].imshow(mTagBFP2_L2_raw, vmax = np.percentile(mTagBFP2_L2_raw[:], 99.5), cmap='gray')\n",
        "ax[0,0].set(title = 'mTagBFP2 L2 raw')\n",
        "ax[0,1].imshow(mTagBFP2_L3_raw, vmax = np.percentile(mTagBFP2_L3_raw[:], 99.5), cmap='gray')\n",
        "ax[0,1].set(title = 'mTagBFP2 L3 raw')\n",
        "ax[1,0].imshow(mTagBFP2_L2_processed, vmax = np.percentile(mTagBFP2_L2_processed[:], 99.5), cmap='gray')\n",
        "ax[1,0].set(title = 'mTagBFP2 L2 processed')\n",
        "ax[1,1].imshow(mTagBFP2_L3_processed, vmax = np.percentile(mTagBFP2_L3_processed[:], 99.5), cmap='gray')\n",
        "ax[1,1].set(title = 'mTagBFP2 L3 processed');"
      ],
      "id": "c39f577e"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "5989a35f"
      },
      "outputs": [],
      "source": [
        "# view content of whole_window_mScarlet_img_2p\n",
        "nwb_retino.processing['ophys']['whole_window_mScarlet_img_2p']"
      ],
      "id": "5989a35f"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "e373dd30"
      },
      "outputs": [],
      "source": [
        "# extract images for mScarlet labeling\n",
        "mScarlet_L2_raw = nwb_retino.processing['ophys']['whole_window_mScarlet_img_2p']['raw_whole_window_mScarlet_img_L2'].data[:]\n",
        "mScarlet_L3_raw = nwb_retino.processing['ophys']['whole_window_mScarlet_img_2p']['raw_whole_window_mScarlet_img_L3'].data[:]\n",
        "mScarlet_L5_raw = nwb_retino.processing['ophys']['whole_window_mScarlet_img_2p']['raw_whole_window_mScarlet_img_L5'].data[:]\n",
        "mScarlet_L2_processed = nwb_retino.processing['ophys']['whole_window_mScarlet_img_2p']['processed_whole_window_mScarlet_img_L2'].data[:]\n",
        "mScarlet_L3_processed = nwb_retino.processing['ophys']['whole_window_mScarlet_img_2p']['processed_whole_window_mScarlet_img_L3'].data[:]\n",
        "mScarlet_L5_processed = nwb_retino.processing['ophys']['whole_window_mScarlet_img_2p']['processed_whole_window_mScarlet_img_L5'].data[:]"
      ],
      "id": "e373dd30"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "2c99bb75"
      },
      "outputs": [],
      "source": [
        "# visualize mTagBFP2 labeling images\n",
        "fig, ax = plt.subplots(2,3, figsize=(12,8))\n",
        "ax[0,0].imshow(mScarlet_L2_raw, vmax = np.percentile(mScarlet_L2_raw[:], 99.5), cmap='gray')\n",
        "ax[0,0].set(title = 'mScarlet L2 raw')\n",
        "ax[0,1].imshow(mScarlet_L3_raw, vmax = np.percentile(mScarlet_L3_raw[:], 99.5), cmap='gray')\n",
        "ax[0,1].set(title = 'mScarlet L3 raw')\n",
        "ax[0,2].imshow(mScarlet_L5_raw, vmax = np.percentile(mScarlet_L5_raw[:], 99.5), cmap='gray')\n",
        "ax[0,2].set(title = 'mScarlet L5 raw')\n",
        "ax[1,0].imshow(mScarlet_L2_processed, vmax = np.percentile(mScarlet_L2_processed[:], 99.5), cmap='gray')\n",
        "ax[1,0].set(title = 'mScarlet L2 processed')\n",
        "ax[1,1].imshow(mScarlet_L3_processed, vmax = np.percentile(mScarlet_L3_processed[:], 99.5), cmap='gray')\n",
        "ax[1,1].set(title = 'mScarlet L3 processed')\n",
        "ax[1,2].imshow(mScarlet_L5_processed, vmax = np.percentile(mScarlet_L5_processed[:], 99.5), cmap='gray')\n",
        "ax[1,2].set(title = 'mScarlet L5 processed');"
      ],
      "id": "2c99bb75"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "a178c833"
      },
      "source": [
        "###  Widefield retinotopy\n",
        "\n",
        "The `retinotopy` processing module contains the results of the widefield images of this mouse, including:\n",
        "- **axis_1_phase_map_altitude**: axis 1 (altitude) phase map in degree\n",
        "- **axis_2_phase_map_azimuth**: axis 2 (azimuth) phase map in degree\n",
        "- **field_sign_map**: the field sign map\n",
        "- **power_map_avg**: average FFT power map (for the stimuli of 4 cordinal directions)\n",
        "- **widefield_vessel_image**: structural image of vessel pattern at brain surface for the whole window; x axis is ~lateral (left) to medial (right), y axis is ~posterior to anterior, same as the \"processed\" vessel image and AAVretro label images under 2P microscope."
      ],
      "id": "a178c833"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "dab1e8cd"
      },
      "outputs": [],
      "source": [
        "# view content of widefield retinotopy\n",
        "nwb_retino.processing['retinotopy']['widefield_imaging']"
      ],
      "id": "dab1e8cd"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "b5be741e"
      },
      "outputs": [],
      "source": [
        "# extract widefield imaging results\n",
        "axis_1_phase_map_altitude = nwb_retino.processing['retinotopy']['widefield_imaging']['axis_1_phase_map_altitude'].data[:]\n",
        "axis_2_phase_map_azimuth = nwb_retino.processing['retinotopy']['widefield_imaging']['axis_2_phase_map_azimuth'].data[:]\n",
        "field_sign_map = nwb_retino.processing['retinotopy']['widefield_imaging']['field_sign_map'].data[:]\n",
        "power_map_avg = nwb_retino.processing['retinotopy']['widefield_imaging']['power_map_avg'].data[:]\n",
        "widefield_vessel_image = nwb_retino.processing['retinotopy']['widefield_imaging']['widefield_vessel_image'].data[:]"
      ],
      "id": "b5be741e"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "11322522"
      },
      "outputs": [],
      "source": [
        "# visualize vessel images\n",
        "fig, ax = plt.subplots(1,2, figsize=(8,4))\n",
        "ax[0].imshow(widefield_vessel_image, vmax = np.percentile(widefield_vessel_image[:], 99.5), cmap='gray')\n",
        "ax[0].set(title = 'Widefield vessel image')\n",
        "ax[1].imshow(vessel_processed, vmax = np.percentile(vessel_processed[:], 99.5), cmap='gray')\n",
        "ax[1].set(title = 'Processed 2P vessel image');"
      ],
      "id": "11322522"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "c4e2f548"
      },
      "source": [
        "### Mouse-specific lab meta data\n",
        "\n",
        "You can find other information about this mouse in the mouse-specific `lab_meta_data`.\n",
        "\n",
        "In **AAVretroInjSite**, you can find where the AAVretro for mScarlet and mTagBFP2 were injected to for this mice. This is the same as `lab_meta_data` for individual sessions.\n",
        "\n",
        "In **Registration**, you can find information about registration of the cranial window to Allen CCF.\n",
        "- **window_center_ml_ccf_mm**: cranial window center location along ML axis in CCF (mm from bregma)\n",
        "- **window_center_AP_ccf_mm**: cranial window center location along AP axis in CCF (mm from bregma)\n",
        "- **vessel_2p_center_ml_ccf_mm**: center location of the processed whole window vessel 2P image along ML axis in CCF (mm from bregma)\n",
        "- **vessel_2p_center_ap_ccf_mm**: center location of the processed whole window vessel 2P image along AP axis in CCF (mm from bregma)\n",
        "- **rs_to_ccf_mm_transformation_matrix**: transformation matrix for transforming the processed whole window vessel 2P image to CCF coordinate (mm from bregma)\n",
        "- **wf_to_rs_transformation_matrix**: transformation matrix for transforming widefield images to the processed whole window vessel 2P image"
      ],
      "id": "c4e2f548"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "296860fd"
      },
      "outputs": [],
      "source": [
        "# view mouse-specific lab meta data\n",
        "nwb_retino.lab_meta_data['harvey_lab_swac_metadata_mouse']"
      ],
      "id": "296860fd"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "b18901cd"
      },
      "outputs": [],
      "source": [
        "# visualize widefield retinotopy results\n",
        "fig, ax = plt.subplots(2,2, figsize=(8,8))\n",
        "ax[0,0].imshow(axis_1_phase_map_altitude, vmax = np.percentile(axis_1_phase_map_altitude[:], 99.5), cmap='gray')\n",
        "ax[0,0].set(title = 'axis 1 phase map (altitude)')\n",
        "ax[0,1].imshow(axis_2_phase_map_azimuth, vmax = np.percentile(axis_2_phase_map_azimuth[:], 99.5), cmap='gray')\n",
        "ax[0,1].set(title = 'axis 2 phase map (azimuth)')\n",
        "ax[1,0].imshow(field_sign_map, vmin = -np.percentile(np.abs(field_sign_map[:]), 99.5),vmax = np.percentile(np.abs(field_sign_map[:]), 99.5), cmap='bwr')\n",
        "ax[1,0].set(title = 'field sign map')\n",
        "ax[1,1].imshow(power_map_avg, vmax = np.percentile(power_map_avg[:], 99.5), cmap='gray')\n",
        "ax[1,1].set(title = 'average power map');"
      ],
      "id": "b18901cd"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QO0T6OanOzzI"
      },
      "source": [
        "# First model regarding calcium related tests"
      ],
      "id": "QO0T6OanOzzI"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "z-3vR-71O1mA"
      },
      "outputs": [],
      "source": [
        "import os\n",
        "import pickle\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn import preprocessing\n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "from sklearn.tree import DecisionTreeClassifier, plot_tree\n",
        "from sklearn.neural_network import MLPClassifier\n",
        "from sklearn.model_selection import GridSearchCV, cross_val_score\n",
        "from sklearn.metrics import f1_score, precision_score, accuracy_score, recall_score, ConfusionMatrixDisplay, confusion_matrix, classification_report\n",
        "import warnings\n",
        "# Disable all FutureWarnings\n",
        "warnings.filterwarnings(\"ignore\", category=FutureWarning)"
      ],
      "id": "z-3vR-71O1mA"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "58QQFJRYPX0e"
      },
      "outputs": [],
      "source": [
        "#convert categorical labels into numerical labels\n",
        "label_encoder = preprocessing.LabelEncoder()\n",
        "\n",
        "#to get the current directory\n",
        "current_directory = os.getcwd()\n",
        "\n",
        "#loading the dataset\n",
        "try:\n",
        "    data = pd.read_csv(current_directory+'\\\\trial_data.csv')\n",
        "except FileNotFoundError:\n",
        "    data = pd.read_csv(current_directory+'/trial_data.csv')\n",
        "\n",
        "valid_data = data.dropna()\n",
        "valid_data = valid_data.sample(frac=1, random_state=42)\n",
        "# Filter rows where is_vis is True\n",
        "true_data = valid_data[valid_data['is_vis'] == True].head(300)\n",
        "\n",
        "# Filter rows where is_vis is False\n",
        "false_data = valid_data[valid_data['is_vis'] == False].head(300)\n",
        "\n",
        "# Concatenate the true and false data together\n",
        "trial = pd.concat([true_data, false_data])\n",
        "\n",
        "#Encoding Categorical Variables\n",
        "trial[\"is_vis\"] = label_encoder.fit_transform(trial[\"is_vis\"])\n",
        "trial[\"is_ruleA\"] = label_encoder.fit_transform(trial[\"is_ruleA\"])\n",
        "trial[\"is_switch\"] = label_encoder.fit_transform(trial[\"is_switch\"])\n",
        "trial[\"is_cueB\"] = label_encoder.fit_transform(trial[\"is_cueB\"])\n",
        "trial[\"is_choL\"] = label_encoder.fit_transform(trial[\"is_choL\"])\n",
        "trial[\"is_correct\"] = label_encoder.fit_transform(trial[\"is_correct\"])\n",
        "\n",
        "#separating the target variable\n",
        "trial_labels = trial[\"is_correct\"]\n",
        "#trial = trial.drop(\"is_correct\", axis=1) #.drop to remove column"
      ],
      "id": "58QQFJRYPX0e"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "--UqHQINW-om"
      },
      "outputs": [],
      "source": [
        "# Play around and see what happens to the models when you change the test-training ratios\n",
        "train_size = 0.8\n",
        "test_size = 1 - train_size\n",
        "assert train_size + test_size == 1\n",
        "\n",
        "trial_features = ['is_choL', 'is_cueB', 'is_ruleA', 'is_vis', 'is_switch']\n",
        "trial = trial[trial_features]\n",
        "\n",
        "# train_test_split this is a method that scikit-learn provides in the library\n",
        "trial_train, trial_test, labels_train, labels_test = train_test_split(trial,trial_labels , train_size=1, random_state=42)"
      ],
      "id": "--UqHQINW-om"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "rt6NALSTY6FP"
      },
      "outputs": [],
      "source": [
        "def metrics(true_y,prediction_y):\n",
        "    print(f\"Accuracy: {accuracy_score(y_true=true_y,y_pred= prediction_y)}\")\n",
        "    print(f\"Precision (Weighted): {precision_score(y_true=true_y,y_pred= prediction_y, average='weighted')}\")\n",
        "    print(f\"Recall (Weighted): {recall_score(y_true=true_y,y_pred= prediction_y, average='weighted')}\")\n",
        "    print(f\"F1 Score (Weighted): {f1_score(y_true=true_y,y_pred= prediction_y,average='weighted')}\")\n",
        "    print(classification_report(y_true=true_y,y_pred= prediction_y, zero_division='warn'))\n",
        "    confusion_matrix(y_true=true_y,y_pred= prediction_y)\n",
        "\n",
        "def compareModels(true_y,prediction_y):\n",
        "    return [accuracy_score(y_true=true_y,y_pred= prediction_y),\n",
        "    precision_score(y_true=true_y,y_pred= prediction_y, average='weighted'),\n",
        "    recall_score(y_true=true_y,y_pred= prediction_y, average='weighted'),\n",
        "    f1_score(y_true=true_y,y_pred= prediction_y,average='weighted')]"
      ],
      "id": "rt6NALSTY6FP"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "vE9f_vOZdApP"
      },
      "outputs": [],
      "source": [
        "import os\n",
        "os.makedirs('models', exist_ok=True)"
      ],
      "id": "vE9f_vOZdApP"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "_9_MJ_TDZA8O"
      },
      "outputs": [],
      "source": [
        "#Bayes\n",
        "classifier = MultinomialNB()\n",
        "if (os.path.exists('models/MNB1_model.pickle')):\n",
        "    pickle_in = open('models/MNB1_model.pickle','rb')\n",
        "    MNBclf = pickle.load(pickle_in)\n",
        "    print(\"pickle file used\")\n",
        "else:\n",
        "    classifier = classifier.fit(trial_train, labels_train)\n",
        "    with open('models/MNB1_model.pickle','wb') as f:\n",
        "        pickle.dump(classifier, f)\n",
        "\n",
        "    pickle_in = open('models/MNB1_model.pickle','rb')\n",
        "    MNBclf = pickle.load(pickle_in)\n",
        "pickle_in.close()"
      ],
      "id": "_9_MJ_TDZA8O"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "G6QMu-cHZuVf"
      },
      "outputs": [],
      "source": [
        "labels_trial_predict = MNBclf.predict(trial_test)\n",
        "\n",
        "metrics(labels_test, labels_trial_predict)\n",
        "cmp = ConfusionMatrixDisplay(confusion_matrix(y_true=labels_test, y_pred= labels_trial_predict))\n",
        "fig, ax = plt.subplots(figsize=(10,10))\n",
        "cmp.plot(ax=ax,cmap='magma')"
      ],
      "id": "G6QMu-cHZuVf"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "go5b8-HRaE26"
      },
      "outputs": [],
      "source": [
        "classifier = DecisionTreeClassifier(criterion=\"entropy\") # default measure of chaos is gini in scikit-learn\n",
        "if (os.path.exists('models/DT1_model.pickle')):\n",
        "    pickle_in = open('models/DT1_model.pickle','rb')\n",
        "    DTclf = pickle.load(pickle_in)\n",
        "    print(\"pickle file used\")\n",
        "else:\n",
        "    classifier = classifier.fit(trial_train, labels_train)\n",
        "    with open('models/DT1_model.pickle','wb') as f:\n",
        "            pickle.dump(classifier, f)\n",
        "    pickle_in = open('models/DT1_model.pickle','rb')\n",
        "    DTclf = pickle.load(pickle_in)\n",
        "pickle_in.close()"
      ],
      "id": "go5b8-HRaE26"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "QCVVX8bYaHsx"
      },
      "outputs": [],
      "source": [
        "labels_trial_predict = DTclf.predict(trial_test)\n",
        "\n",
        "metrics(labels_test, labels_trial_predict)\n",
        "cmp = ConfusionMatrixDisplay(confusion_matrix(y_true=labels_test, y_pred= labels_trial_predict))\n",
        "fig, ax = plt.subplots(figsize=(10,10))\n",
        "cmp.plot(ax=ax,cmap='magma')"
      ],
      "id": "QCVVX8bYaHsx"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Sm0bCXCPaJvP"
      },
      "outputs": [],
      "source": [
        "# Discretize is_vis feature into two bins: <= 0.5 and > 0.5\n",
        "trial['is_vis_bin'] = pd.cut(trial['is_vis'], bins=[-np.inf, 0.5, np.inf], labels=[0, 1], include_lowest=True, right=False)  # Use numerical labels\n",
        "\n",
        "# Define features and target variable\n",
        "trial_features = ['is_vis_bin', 'is_ruleA', 'is_switch', 'is_cueB', 'is_choL']\n",
        "X = trial[trial_features]\n",
        "y = trial_labels\n",
        "\n",
        "# Train the decision tree model\n",
        "classifier = DecisionTreeClassifier(criterion=\"entropy\")\n",
        "DTclf = classifier.fit(X, y)\n",
        "\n",
        "# Define class names as strings\n",
        "class_names = ['Incorrect', 'Correct']\n",
        "\n",
        "classifier = DecisionTreeClassifier(criterion=\"entropy\")\n",
        "DTclf = classifier.fit(X, y)\n",
        "\n",
        "# Plot the decision tree\n",
        "plt.figure(figsize=(25, 15))\n",
        "plot_tree(DTclf, feature_names=trial_features, class_names=class_names, fontsize=10)\n",
        "plt.show()\n",
        "\n",
        "# Analyze feature importance\n",
        "feature_importance = DTclf.feature_importances_\n",
        "print(\"Feature Importance:\", feature_importance)\n"
      ],
      "id": "Sm0bCXCPaJvP"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jkY6Ka-MEj_D"
      },
      "source": [
        "# Second Trial regarding widefield retinotopy"
      ],
      "id": "jkY6Ka-MEj_D"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "8SBovGUB4EF4"
      },
      "outputs": [],
      "source": [
        "!pip install dandi\n",
        "!dandi download https://api.dandiarchive.org/api/assets/3362af75-91fc-4e40-8536-aabcb74f8f15/download/"
      ],
      "id": "8SBovGUB4EF4"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Iq5gG0Kz4Cm_"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "from pynwb import NWBHDF5IO\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.model_selection import train_test_split\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Conv1D, Conv2D, MaxPooling2D, Flatten, Dense\n",
        "from tensorflow.keras.callbacks import EarlyStopping\n",
        "from skimage.transform import resize\n",
        "import os\n",
        "\n",
        "# Chemin de mes 2 fichiers\n",
        "train_file_path = './sub-9_ses-mouse-9-session-date-2017-08-11-area-visA-L23-multi-plane-imaging_behavior+ophys.nwb'\n",
        "test_file_path = './sub-9_ses-mouse-9-session-date-2017-08-11-area-visA-L23-multi-plane-imaging_behavior+ophys.nwb'\n",
        "\n",
        "# Code pour extraire les données d'un fichier NWB\n",
        "io_train = NWBHDF5IO(train_file_path, 'r')\n",
        "nwb_train = io_train.read()\n",
        "\n",
        "io_test = NWBHDF5IO(test_file_path, 'r')\n",
        "nwb_test = io_test.read()\n",
        "\n",
        "# Extraction des données de neurones et des étiquettes de position de la souris\n",
        "deconv_0 = nwb_train.processing['ophys']['deconvolved_activity_plane_0'].data[:]\n",
        "\n",
        "# Extraction des données de neurones et des étiquettes de position de la souris\n",
        "ml_train = nwb_train.processing['ophys']['ImageSegmentation']['PlaneSegmentation_0'].to_dataframe()['ml'].values\n",
        "ap_train = nwb_train.processing['ophys']['ImageSegmentation']['PlaneSegmentation_0'].to_dataframe()['ap'].values\n",
        "depth_train = nwb_train.processing['ophys']['ImageSegmentation']['PlaneSegmentation_0'].to_dataframe()['depth'].values\n",
        "area_train = nwb_train.processing['ophys']['ImageSegmentation']['PlaneSegmentation_0'].to_dataframe()['area'].values\n",
        "deconv_0 = nwb_train.processing['ophys']['deconvolved_activity_plane_0'].data[:]\n",
        "\n",
        "ml_test = nwb_test.processing['ophys']['ImageSegmentation']['PlaneSegmentation_0'].to_dataframe()['ml'].values\n",
        "ap_test = nwb_test.processing['ophys']['ImageSegmentation']['PlaneSegmentation_0'].to_dataframe()['ap'].values\n",
        "depth_test = nwb_test.processing['ophys']['ImageSegmentation']['PlaneSegmentation_0'].to_dataframe()['depth'].values\n",
        "area_test = nwb_test.processing['ophys']['ImageSegmentation']['PlaneSegmentation_0'].to_dataframe()['area'].values\n",
        "\n",
        "# convert the first PlaneSegmentation into dataframe and examine the information\n",
        "planeSeg_0_df = nwb_train.processing['ophys']['ImageSegmentation']['PlaneSegmentation_0'].to_dataframe()\n",
        "planeSeg_0_df.head(20)\n",
        "# extract information for these neurons in plane 0\n",
        "ml_0 = planeSeg_0_df['ml'].values\n",
        "ap_0 = planeSeg_0_df['ap'].values\n",
        "depth_0 = planeSeg_0_df['depth'].values\n",
        "area_0 = planeSeg_0_df['area'].values\n",
        "mTagBFP2_0 = planeSeg_0_df['mTagBFP2'].values\n",
        "mScarlet_0 = planeSeg_0_df['mScarlet'].values\n",
        "\n",
        "# compute number of neurons in this plane\n",
        "n_neurons_0 = ml_0.shape[0]\n",
        "\n",
        "# convert trial table into dataframe and examine it\n",
        "trial_df = nwb_train.trials.to_dataframe()\n",
        "trial_df.head()\n",
        "# extract basic trial information\n",
        "cueB = trial_df['is_cueB'].values.astype(int)\n",
        "choL = trial_df['is_choL'].values.astype(int)\n",
        "vis = trial_df['is_vis'].values.astype(int)\n",
        "ruleA = trial_df['is_ruleA'].values.astype(int)\n",
        "correctness = trial_df['is_correct'].values.astype(int)\n",
        "switches = np.argwhere(trial_df['is_switch'].values).flatten()\n",
        "# get number of trials\n",
        "nTrials = cueB.shape[0]\n",
        "\n",
        "# compute moving average of correct for every 7 trials\n",
        "corr_moving_avg = np.convolve(correctness, np.ones(7,)/7,mode='same')\n",
        "# extract frame-aligned behavioral variables\n",
        "imaging_frame_timestamps = nwb_train.processing['behavior']['frame_aligned_position']['frame_aligned_forward_and_lateral_position'].timestamps[:]\n",
        "frame_aligned_position = nwb_train.processing['behavior']['frame_aligned_position']['frame_aligned_forward_and_lateral_position'].data[:]\n",
        "frame_aligned_velocity = nwb_train.processing['behavior']['frame_aligned_velocity']['frame_aligned_pitch_roll_yaw_velocity'].data[:]\n",
        "frame_aligned_trial_number = nwb_train.processing['behavior']['frame_aligned_trial_number'].data[:]\n",
        "frame_aligned_time_from_choice_point = nwb_train.processing['behavior']['frame_aligned_time_from_choice_point'].data[:]\n",
        "plane_idx_for_imaging_frames = nwb_train.processing['behavior']['plane_idx_for_imaging_frames'].data[:]\n",
        "# identify valid trials (trials that have corresponding imaging frames)\n",
        "validTrials = np.unique(frame_aligned_trial_number[~np.isnan(frame_aligned_trial_number)]).astype(int)\n",
        "nValidTrials = validTrials.shape[0]\n",
        "## extract important task parameters from lab meta data\n",
        "# maze configuration\n",
        "meter_per_virmen_unit = nwb_train.lab_meta_data['harvey_lab_swac_metadata_session'].TaskParam__meter_per_virmen_unit # 1 Virmen unit in the maze ~= 0.01 m\n",
        "stem_length = nwb_train.lab_meta_data['harvey_lab_swac_metadata_session'].TaskParam__maze_stem_length*meter_per_virmen_unit\n",
        "arm_length = nwb_train.lab_meta_data['harvey_lab_swac_metadata_session'].TaskParam__maze_arm_length*meter_per_virmen_unit\n",
        "\n",
        "# the length of the cue delay before the maze arms (sueful if you're analyzing visual response to the cue)\n",
        "cue_delay_length = nwb_train.lab_meta_data['harvey_lab_swac_metadata_session'].TaskParam__cue_delay_length*meter_per_virmen_unit\n",
        "\n",
        "# the length of the feedback delay in seconds\n",
        "feedback_delay = nwb_train.lab_meta_data['harvey_lab_swac_metadata_session'].TaskParam__feedback_delay_sec\n",
        "\n",
        "# the length of the reward delay in seconds (equals length of the feedback delay + feedback period)\n",
        "reward_delay = nwb_train.lab_meta_data['harvey_lab_swac_metadata_session'].TaskParam__reward_delay_sec\n",
        "\n",
        "# the lenght of ITI for correct and incorrect trials\n",
        "iti_correct = nwb_train.lab_meta_data['harvey_lab_swac_metadata_session'].TaskParam__iti_correct_sec\n",
        "iti_incorrect = nwb_train.lab_meta_data['harvey_lab_swac_metadata_session'].TaskParam__iti_incorrect_sec\n",
        "# set up position bins\n",
        "pos_half_width = 0.1\n",
        "maze_length = (stem_length + arm_length)\n",
        "pos_bins = np.arange(0.1, maze_length + 0.05, pos_half_width)\n",
        "# position and temporal binning for deconvolved activity\n",
        "activity_pos = np.full((nValidTrials, pos_bins.shape[0], n_neurons_0), np.NaN)\n",
        "\n",
        "# set up time bins\n",
        "tm_half_width = 1/3\n",
        "feedback_iti_corr_length = reward_delay + iti_correct\n",
        "tm_bins = np.arange(0, feedback_iti_corr_length + 0.1, tm_half_width)\n",
        "\n",
        "# prelocate\n",
        "vel_pos = np.full((nValidTrials, pos_bins.shape[0]), np.NaN)\n",
        "vel_tm = np.full((nValidTrials, tm_bins.shape[0]), np.NaN)\n",
        "\n",
        "activity_tm = np.full((nValidTrials, tm_bins.shape[0], n_neurons_0), np.NaN)\n",
        "\n",
        "# use plane_idx_for_imaging_frames to identify behavioral variables corresponding to this imaging plane\n",
        "nPlane = 0\n",
        "frame_idx_0 = plane_idx_for_imaging_frames == nPlane\n",
        "trial_number_0 = frame_aligned_trial_number[frame_idx_0]\n",
        "# get forward and lateral positions\n",
        "posF = frame_aligned_position[:,0]\n",
        "posF_0 = posF[frame_idx_0]\n",
        "time_from_choice_point_0 = frame_aligned_time_from_choice_point[frame_idx_0]"
      ],
      "id": "Iq5gG0Kz4Cm_"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "nYcV_55xJERM",
        "outputId": "6a8a7e99-ec0e-4029-9722-8474cdfafab8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mean Squared Errors: [0.5146089922408188, 0.5363854661106344, 0.5313592058399437, 0.5300448222891163, 0.46348066524304227, 0.5288218196757746, 0.5337691233852795, 0.5168171323551644, 0.35034480455456496, 0.5399565320952782, 0.5287456633554376, 0.5348607475659277, 0.5433759366931423, 0.5376532111967276, 0.5340260289595543, 0.5457266282729771, 0.4624498626073219, 0.5207354854926669, 0.5374400811351324, 0.4838431300214497, 0.5361617305425914, 0.541174304055277, 0.5388417023921026, 0.5356023522849102, 0.5361561011655354, 0.5370372187024219, 0.53620613699492, 0.5361119633002669, 0.5346007355738411, 0.5327746962039703, 0.5120306598481295, 0.57123957025688, 0.5344650519263426, 0.5340429545963593, 0.5270013228782278, 0.536273495746461, 0.5322408879580478, 0.5273903030404187, 0.5199261305597442, 0.5394697606372901, 0.5392353328083448, 0.5415811467509065, 0.5382007576324501, 0.5286190866843178, 0.5364509024823867, 0.5393486589512833, 0.5362861658150964, 0.5293945986147396, 0.5301498562324298, 0.5142811280412042, 0.5407342400610521, 0.5253176248122349, 0.5338593971919394, 0.5362654862462103, 0.49566156728962285, 0.4961334809198769, 0.5360461187463068, 0.5448030499002001, 0.535566898089565, 0.5458677459643788, 0.536273495746461, 0.5207051010446577, 0.5387973638754797, 0.5184698828787011, 0.5139493327833768, 0.4880316930991103, 0.5328688842207183, 0.5401701537977451, 0.558188492944075, 0.5362938555375582, 0.546166408072257, 0.5366136628335398, 0.5510779956863062, 0.5329378078446597, 0.5070495631176455, 0.532936816635697, 0.5198186493810183, 0.5148577108113318, 0.5336069820502607, 0.5400109556934034, 0.5382876075358551, 0.5540549337186474, 0.5295313442371326, 0.536273495746461, 0.5450829423614041, 0.5435408347630167, 0.5354053939320085, 0.5398710244210466, 0.4809349868726077, 0.5396408594738057, 0.5362226888463486, 0.5374830304728743, 0.5342250490661125, 0.5340904583669432, 0.5326139223492575, 0.536324603789647, 0.529714073467219, 0.5282901604302136, 0.5361167358515123, 0.5387014329206549, 0.4539146162615066, 0.536277104000844, 0.5364373397554973, 0.4598659072930122, 0.5233905764784328, 0.478854114229189, 0.523582715558065, 0.5261751688013895, 0.532315261907507, 0.5303758368993675, 0.4805463655333059, 0.5452387908355224, 0.5773947872324958, 0.5264520793241461, 0.536037672225358, 0.5332927628453363, 0.5286012287316628, 0.5323329816897091, 0.536273495746461, 0.5369244366559555, 0.5415405754001712, 0.5344351321640117, 0.5115602657690467, 0.5366241729048367, 0.5364184589237067, 0.536273495746461, 0.536273495746461, 0.5218503866255102, 0.5411767500680108, 0.5396642295628803, 0.5135737439842445, 0.5398207241505582, 0.535882264633052, 0.5083610243734265, 0.536273495746461, 0.5083980500445591, 0.5333489103888537, 0.534961043596928, 0.5191100114053973, 0.529854521709075, 0.5462505922239004, 0.5321959446710507, 0.49973990899729853, 0.5356358062672827, 0.5349109582579455, 0.5243425028124541, 0.5371383490700808, 0.5597980077094572, 0.5356792748402753, 0.45267213145736607, 0.5380100907521858, 0.5215575877956491, 0.5336469238457507, 0.5485168143062122, 0.5211156564919703, 0.5353929719541479, 0.5357673679147092, 0.5362487069500477, 0.5386252254165159, 0.4878919262404713, 0.5320910253863305, 0.5362204410744968, 0.5427952551578515, 0.5332694368867275, 0.5536385672007953, 0.5354408925241331, 0.5239159523584609, 0.5210831020650135, 0.5236906152321461, 0.5390077844987349, 0.536273495746461, 0.5169693319753146, 0.5449017918001235, 0.5339196402524367, 0.536534691282648, 0.5324167207960369, 0.5219725895273012, 0.5239845626786626, 0.5282671289211661, 0.5317859518899783, 0.5401323867282295, 0.5369249116518343, 0.5363514079304514, 0.48416074566067807, 0.5410871163780773, 0.5283228323899389, 0.5345394452900412, 0.5197819611722992, 0.5160985176990126, 0.5362803631254879, 0.533520107327069, 0.5370963720092993, 0.528488766924497, 0.5025747410790331, 0.5211897711372206, 0.5364264197922586, 0.5362369917131431, 0.5046605679503886, 0.5349520560861286, 0.5327814864429568, 0.5502823650936997, 0.5320138346967747, 0.5374965092594788, 0.5267866248859814, 0.5205674144427364, 0.5176521620278685, 0.5198806942624667, 0.5365402911454267, 0.5369783102880936, 0.5335159201702089, 0.5188099457360791, 0.5427174376707792, 0.5362676418247411, 0.5400505446474361, 0.527650788005471, 0.5358206043376359, 0.5346467912282229, 0.54471823279907, 0.5342140990697859, 0.5336759365575687, 0.5405930347508434, 0.527337552994816, 0.5226574903956556, 0.5236345485841811, 0.5284578189776943, 0.5365775335279295, 0.5314167563129027, 0.5337694084975814, 0.5075647407526961, 0.5050493473566291, 0.5355958811503949, 0.5370009312027023, 0.5410628918596034, 0.5280453548213464, 0.5009224785663859, 0.5376649589733604, 0.5239584837499442, 0.5349626077425155, 0.5347959192375022, 0.5250102741324084, 0.5366566938045978, 0.5336589994053449, 0.536318812403635, 0.5337301172467661, 0.509249078407916, 0.5258800737033698, 0.5363465986837992, 0.5156818660665387, 0.5522745289223475, 0.5118225222539342, 0.5365338736246323, 0.5378704934658997, 0.5375813476331707, 0.5299569723928274, 0.4988645670463657, 0.5281308788720949, 0.5300046724887565, 0.5355857170638068, 0.5363744639186913, 0.5328189965837519, 0.5270395145594506, 0.518905920583865, 0.53652995869422, 0.5110404229495046, 0.5219252731759492]\n"
          ]
        }
      ],
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.linear_model import LinearRegression\n",
        "from sklearn.metrics import mean_squared_error\n",
        "import numpy as np\n",
        "\n",
        "# Préparation des données\n",
        "fill_value = n_neurons_0  # Valeur de remplissage pour np.full()\n",
        "deconv_0_expanded = np.expand_dims(deconv_0, axis=1)\n",
        "activity_pos_selected = activity_pos[:, :nValidTrials, :]\n",
        "\n",
        "# Création et entraînement du modèle pour chaque neurone\n",
        "models = []\n",
        "mses = []\n",
        "for i in range(n_neurons_0):\n",
        "    X = np.concatenate((deconv_0_expanded[:nValidTrials, :, i], np.full((nValidTrials, 1), fill_value), activity_pos_selected[:, :, i]), axis=1)\n",
        "    y = posF_0[:nValidTrials]\n",
        "\n",
        "    # Remplacement des valeurs NaN par zéro\n",
        "    X = np.nan_to_num(X)\n",
        "    y = np.nan_to_num(y)\n",
        "\n",
        "    # Division des données en ensembles d'entraînement et de test\n",
        "    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)\n",
        "\n",
        "    # Création et entraînement du modèle\n",
        "    model = LinearRegression()\n",
        "    model.fit(X_train, y_train)\n",
        "    models.append(model)\n",
        "\n",
        "    # Prédiction sur l'ensemble de test\n",
        "    y_pred = model.predict(X_test)\n",
        "\n",
        "    # Évaluation du modèle\n",
        "    mse = mean_squared_error(y_test, y_pred)\n",
        "    mses.append(mse)\n",
        "\n",
        "print(f\"Mean Squared Errors: {mses}\")"
      ],
      "id": "nYcV_55xJERM"
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.11.0"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 5
}