notebooks/monkey/08_MONKEY_LatentAnalysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Figure 6 Decoding Performance vs. latent uncertainty\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import gc\n",
    "import os\n",
    "import pickle\n",
    "from datetime import datetime\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import torch\n",
    "import scipy.stats as stats\n",
    "\n",
    "from maskedvae.plotting.plotting import (make_col_dict_monkey,get_title_dict,cm2inch)\n",
    "from maskedvae.model.networks import apply_mask\n",
    "from maskedvae.utils.evaluate import (\n",
    "    infer_latents_monkey,\n",
    "    train_on_all_obs,\n",
    "    predict_and_evaluate,\n",
    ")\n",
    "from maskedvae.utils.utils import gpu, scale_datasets, copy_and_shuffle\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "fontsize=10\n",
    "plt.rcParams.update({'font.size': fontsize})\n",
    "plt.rcParams.update({'axes.spines.right': False})\n",
    "plt.rcParams.update({'axes.spines.top': False})\n",
    "plt.rcParams.update({'xtick.labelsize': fontsize})\n",
    "plt.rcParams.update({'ytick.labelsize': fontsize})\n",
    "plt.rcParams.update({'axes.titlesize': fontsize})\n",
    "plt.rcParams.update({'axes.labelsize': fontsize})\n",
    "plt.rcParams.update({'legend.fontsize': fontsize})\n",
    "  \n",
    "# specify the directory here where runs are stored\n",
    "temp_dir = \"../../runs/monkey/\"\n",
    "\n",
    "maskedloss=True\n",
    "new_date=True\n",
    "\n",
    "col_dict = make_col_dict_monkey()\n",
    "title_dict = get_title_dict()\n",
    "\n",
    "# reload the developing code \n",
    "%load_ext autoreload\n",
    "%autoreload 2\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "run_files = \"./R_all_obs_40_mixed_extended.txt\"\n",
    "# load run_dirs from a text file\n",
    "with open(run_files, \"r\") as f:\n",
    "    run_dirs = f.readlines()\n",
    "run_dirs = [x.strip() for x in run_dirs]\n",
    "\n",
    "dimwise = True\n",
    "cross_loss = False\n",
    "beta_val = 1\n",
    "nll_beta = 0.3\n",
    "\n",
    "selected_runs_masked = []\n",
    "selected_runs_all_obs = []\n",
    "latents = \"_lat_40\"\n",
    "full_mask_prob = 0.2\n",
    "for run in run_dirs:\n",
    "\n",
    "    if latents in run:\n",
    "        with open(temp_dir + run + \"/args.pkl\", \"rb\") as f:\n",
    "            args = pickle.load(f)\n",
    "        if len(args.run_params[\"outputs\"]) == 1:\n",
    "            if \"all_obs\" in run:\n",
    "                # all observed naive runs\n",
    "                selected_runs_all_obs.append(run)\n",
    "            else:\n",
    "                # check if 0.2 or 0.5 all observed chance\n",
    "                if args.run_params[\"full_mask_prob\"] == full_mask_prob:\n",
    "                    selected_runs_masked.append(run)\n",
    "\n",
    "\n",
    "run_dirs = selected_runs_masked + selected_runs_all_obs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "count_masked: 10, count_naive: 10\n"
     ]
    }
   ],
   "source": [
    "tag_dim = \"dimwise\" if dimwise else \"no_dimwise\"\n",
    "date_tag = datetime.now().strftime(\"%d%m%y\").replace(\".\", \"_\")\n",
    "tag_cross = \"cross_loss\" if cross_loss else \"no_cross_loss\"\n",
    "tag_cross = date_tag + \"_\" + tag_cross\n",
    "\n",
    "\n",
    "local_dir = \"../../data/monkey/\"\n",
    "\n",
    "# make dires if they dont exist\n",
    "if not os.path.exists(local_dir):\n",
    "    os.makedirs(local_dir)\n",
    "if not os.path.exists(local_dir + \"figures/\"):\n",
    "    os.makedirs(local_dir + \"figures/\")\n",
    "\n",
    "\n",
    "# check how many masked and naive runs are available\n",
    "count_masked = 0\n",
    "count_naive = 0\n",
    "for run in run_dirs:\n",
    "    count_masked += 1 if \"masked\" in run else 0\n",
    "    count_naive += 1 if \"all_obs\" in run else 0\n",
    "print(f\"count_masked: {count_masked}, count_naive: {count_naive}\")\n",
    "\n",
    "if count_naive > count_masked:\n",
    "    run_dirs = run_dirs[: 2 * count_masked]\n",
    "\n",
    "\n",
    "# get the modeltags naive and run names\n",
    "model_tags = []\n",
    "run_names = []\n",
    "\n",
    "for q, run in enumerate(run_dirs):\n",
    "    run_names.append(run)\n",
    "    if \"all_obs\" in run:\n",
    "        model_tags.append(\"naive\")\n",
    "    elif \"masked\" in run:\n",
    "        model_tags.append(\"masked\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluate the masked latents for all runs loading each model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "t_min = 100\n",
    "t_max = 300\n",
    "lat_mask = None\n",
    "inp_mask = (\"xa_s\",)\n",
    "test_masks = [\n",
    "    \"all_obs\",\n",
    "    \"xa_m_5\",\n",
    "    \"xa_m_20\",\n",
    "    \"xa_m_50\",\n",
    "    \"xa_m_100\",\n",
    "    \"xa_m_150\",\n",
    "    \"xa_m_200\",\n",
    "]\n",
    "\n",
    "# specify the colors for the different masks\n",
    "reds = plt.get_cmap(\"Reds\")(np.linspace(0.3, 0.95, len(test_masks)))\n",
    "blues = plt.get_cmap(\"Blues\")(np.linspace(0.3, 0.95, len(test_masks)))\n",
    "\n",
    "# load the data\n",
    "data_dir = \"../../data/monkey/\"\n",
    "with open(data_dir + \"/PD_train.pkl\", \"rb\") as f:\n",
    "    PD_train = pickle.load(f)\n",
    "with open(data_dir + \"/PD_valid.pkl\", \"rb\") as f:\n",
    "    PD_valid = pickle.load(f)\n",
    "with open(data_dir + \"/PD_test.pkl\", \"rb\") as f:\n",
    "    PD_test = pickle.load(f)\n",
    "\n",
    "# Loop over q values (assuming you have a list `run_dirs` and `model_tags`)\n",
    "spikes_matrices = []\n",
    "latent_means_masked = []\n",
    "latent_stds_masked = []\n",
    "latent_means_naive = []\n",
    "latent_stds_naive = []\n",
    "# loop over the runs and load the model and pass the masked test data through the model\n",
    "for q in range(len(run_dirs)):\n",
    "    exp_dir = temp_dir + run_dirs[q] + \"/\"\n",
    "\n",
    "    with open(exp_dir + \"model.pkl\", \"rb\") as f:\n",
    "        model = torch.load(f)\n",
    "\n",
    "    session_data = PD_test.get_session(\n",
    "        0, to_gpu=True, t_slice=model.eval_params[\"t_slice_test\"]\n",
    "    )\n",
    "\n",
    "    if lat_mask is None:\n",
    "        lat_mask = gpu(torch.ones(model.net.n_latents))\n",
    "\n",
    "    # Run network\n",
    "    latent_means = []\n",
    "    latent_stds = []\n",
    "    t = 0\n",
    "    # standardize the the behavioral data - here no output behavior is used\n",
    "    for k in model.net.scaling:\n",
    "        if model.net.ifdimwise_scaling:\n",
    "            for ii in range(session_data[k].shape[1]):\n",
    "                session_data[k][:, ii] = model.net.dimwise_scaling[k + str(ii)][t][\n",
    "                    0\n",
    "                ] * (\n",
    "                    session_data[k][:, ii]\n",
    "                    - model.net.dimwise_scaling[k + str(ii)][t][1]\n",
    "                )\n",
    "        else:\n",
    "            session_data[k] = model.net.scaling[k][t][0] * (\n",
    "                session_data[k] - model.net.scaling[k][t][1]\n",
    "            )  # Scale traces for training\n",
    "        print(k, \"Scaled... \")\n",
    "\n",
    "    model_tag = \"naive\" if model.full_mask == 1 else \"masked\"\n",
    "    # select different test masks\n",
    "\n",
    "    for mask_key in test_masks:\n",
    "        # if the mask has not been generated for the model, skip\n",
    "        if mask_key not in model.masks.keys() and mask_key != \"all_obs\":\n",
    "            print(mask_key + \" not in model.masks.keys() or mask_key != all_obs\")\n",
    "            continue\n",
    "        # for all masks that are part of the model pass the mask as train_mask (multiplies the mask with the input)\n",
    "        if mask_key != \"all_obs\":\n",
    "            print(mask_key, \"Masked... \")\n",
    "            network_outputs = model.net.encode(\n",
    "                session_data,\n",
    "                train_mask=model.masks[mask_key][: model.net.n_inputs],\n",
    "                inp_mask=inp_mask,\n",
    "            )\n",
    "            masked_spikes = apply_mask(\n",
    "                model,\n",
    "                session_data,\n",
    "                train_mask=model.masks[mask_key][: model.net.n_inputs],\n",
    "                inp_mask=inp_mask,\n",
    "            )\n",
    "            if q == 0:\n",
    "                spikes_matrices.append(masked_spikes[0].detach().cpu().numpy())\n",
    "        else:  # for all observed train_mask is None and masking step is skipped\n",
    "            print(\"all_observed\")\n",
    "            network_outputs = model.net.encode(session_data, inp_mask=inp_mask)\n",
    "            masked_spikes = apply_mask(model, session_data, inp_mask=inp_mask)\n",
    "            if q == 0:\n",
    "                spikes_matrices.append(masked_spikes[0].detach().cpu().numpy())\n",
    "\n",
    "        latent_means.append(network_outputs[\"z_mu\"].detach().cpu().numpy())\n",
    "        latent_stds.append(torch.exp(network_outputs[\"z_lsig\"]).detach().cpu().numpy())\n",
    "    # append the latent means and stds to the respective lists\n",
    "    if model_tag == \"masked\":\n",
    "        latent_means_masked.append(latent_means)\n",
    "        latent_stds_masked.append(latent_stds)\n",
    "    else:\n",
    "        latent_means_naive.append(latent_means)\n",
    "        latent_stds_naive.append(latent_stds)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Figure 6 Panel B & C:\n",
    "### B: Masked neural population activity for different masking level\n",
    "### C: corresponding inferred latents of masked and naive VAEs "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 748.031x393.701 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Define the indices for the test masks you want to plot\n",
    "plot_test_masks = [\n",
    "    \"all_obs\",\n",
    "    \"xa_m_50\",\n",
    "    \"xa_m_100\",\n",
    "    \"xa_m_150\",\n",
    "]\n",
    "\n",
    "masked_vals = [\n",
    "    0,\n",
    "    50,\n",
    "    100,\n",
    "    150,\n",
    "]\n",
    "\n",
    "\n",
    "# Specify time and latent index for the latent_means plot\n",
    "t_min, t_max = 100, 160  # example time window (adjust according to your data)\n",
    "\n",
    "q_masked, q_naive = 3, 3  # example run indices to plot\n",
    "limval = 2.5\n",
    "# list the latent indices sorted by variance highest first\n",
    "sorted_indices_m = np.argsort(\n",
    "    np.std(latent_means_masked[q_masked][0][0, :, t_min:t_max], axis=1)\n",
    ")[::-1]\n",
    "latent_index_m = sorted_indices_m[0]  # example latent index to plot\n",
    "\n",
    "# list the latent indices sorted by variance highest first\n",
    "sorted_indices_n = np.argsort(\n",
    "    np.std(latent_means_naive[q_naive][0][0, :, t_min:t_max], axis=1)\n",
    ")[::-1]\n",
    "latent_index_n = sorted_indices_n[0]  # example latent index to plot\n",
    "\n",
    "# Prepare the figure layout\n",
    "fig, axs = plt.subplots(\n",
    "    3,\n",
    "    4,\n",
    "    figsize=cm2inch(19, 10),\n",
    "    gridspec_kw={\"height_ratios\": [1.5, 1, 1]},\n",
    "    sharey=\"row\",\n",
    ")\n",
    "# make the y axis of the last two rows the same\n",
    "axs[1, 0].set_ylim(-limval, limval)\n",
    "axs[2, 0].set_ylim(-limval, limval)\n",
    "\n",
    "\n",
    "# Iterate over the 4 test masks and plot the corresponding spike trains and latent means\n",
    "for mm, mask in enumerate(plot_test_masks):\n",
    "\n",
    "    # get the index the mask would have in the test_masks list\n",
    "    mask_index = test_masks.index(mask)\n",
    "\n",
    "    # Plot spike trains (top row)\n",
    "    ax_spike = axs[0, mm]\n",
    "    ax_spike.imshow(\n",
    "        spikes_matrices[mask_index][:, t_min:t_max], aspect=\"auto\", cmap=\"Greys\"\n",
    "    )\n",
    "    # get the number of masked neurons and draw a black line to separate them\n",
    "    ax_spike.axhline(y=masked_vals[mm], color=\"black\", linestyle=\"-\")\n",
    "\n",
    "    title = mask.replace(\"_\", \" \").replace(\"xa m\", \"mskd\")\n",
    "    ax_spike.set_title(title)\n",
    "    ax_spike.set_xticks([])\n",
    "\n",
    "    # Plot latent means with std band (bottom row 1)\n",
    "    ax_latent_mean = axs[1, mm]\n",
    "    latent_mean = latent_means_masked[q_masked][mask_index][\n",
    "        0, latent_index_m, t_min:t_max\n",
    "    ]\n",
    "    latent_std = latent_stds_masked[q_masked][mask_index][\n",
    "        0, latent_index_m, t_min:t_max\n",
    "    ]\n",
    "    time = np.arange(t_min, t_max) / model.fps\n",
    "\n",
    "    ax_latent_mean.plot(time, latent_mean, label=mask, color=reds[mask_index])\n",
    "    ax_latent_mean.fill_between(\n",
    "        time,\n",
    "        latent_mean - latent_std,\n",
    "        latent_mean + latent_std,\n",
    "        color=reds[1],\n",
    "        alpha=0.3,\n",
    "    )\n",
    "    ax_latent_mean.plot(\n",
    "        time, latent_mean - latent_std, color=reds[mask_index], alpha=0.5\n",
    "    )\n",
    "    ax_latent_mean.plot(\n",
    "        time, latent_mean + latent_std, color=reds[mask_index], alpha=0.5\n",
    "    )\n",
    "    # ax_latent_mean.set_xlabel('time')\n",
    "    ax_latent_mean.set_xticks([])\n",
    "    # Plot latent means naive\n",
    "    ax_latent_mean = axs[2, mm]\n",
    "    latent_mean = latent_means_naive[q_naive][mask_index][\n",
    "        0, latent_index_n, t_min:t_max\n",
    "    ]\n",
    "    latent_std = latent_stds_naive[q_naive][mask_index][0, latent_index_n, t_min:t_max]\n",
    "    # time = np.arange(t_min, t_max)\n",
    "\n",
    "    ax_latent_mean.plot(time, latent_mean, label=mask, color=blues[mask_index])\n",
    "    ax_latent_mean.fill_between(\n",
    "        time,\n",
    "        latent_mean - latent_std,\n",
    "        latent_mean + latent_std,\n",
    "        color=blues[1],\n",
    "        alpha=0.3,\n",
    "    )\n",
    "    ax_latent_mean.plot(\n",
    "        time, latent_mean - latent_std, color=blues[mask_index], alpha=0.5\n",
    "    )\n",
    "    ax_latent_mean.plot(\n",
    "        time, latent_mean + latent_std, color=blues[mask_index], alpha=0.5\n",
    "    )\n",
    "\n",
    "    ax_latent_mean.set_xlabel(\"time [s]\")\n",
    "\n",
    "axs[0, 0].set_ylabel(\"neuron id\")\n",
    "axs[1, 0].set_ylabel(\"latent masked\")\n",
    "axs[2, 0].set_ylabel(\"latent naive\")\n",
    "for ax in axs.flatten():\n",
    "    ax.spines[\"left\"].set_position((\"outward\", 4))\n",
    "\n",
    "# Adjust layout\n",
    "# plt.tight_layout()\n",
    "plt.savefig(\n",
    "    f\"{local_dir}/figures/latent_mean_std_masked_vs_naive_{q_masked}_{q_naive}.png\",\n",
    "    bbox_inches=\"tight\",\n",
    ")\n",
    "plt.savefig(\n",
    "    f\"{local_dir}/figures/latent_mean_std_masked_vs_naive_{q_masked}_{q_naive}.pdf\",\n",
    "    bbox_inches=\"tight\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Figure 6 Panel D:\n",
    "### Distribution of latent uncertainty for all masking levels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 610.236x78.7402 with 7 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ------------------------------------------------------\n",
    "#\n",
    "# plot the histograms of the latent stds\n",
    "#\n",
    "# ------------------------------------------------------\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots(\n",
    "    1, len(test_masks), figsize=cm2inch(15.5, 2), sharey=True, sharex=True\n",
    ")\n",
    "ax = ax.flatten()\n",
    "bins = np.linspace(0.1, 1.2, 31)\n",
    "t_min, t_max = 50, 10000\n",
    "for mask_index in range(len(test_masks)):\n",
    "    ax[mask_index].hist(\n",
    "        latent_stds_naive[q_naive][mask_index][0, latent_index_n, t_min:t_max],\n",
    "        bins=bins,\n",
    "        color=\"darkblue\",\n",
    "        histtype=\"step\",\n",
    "        lw=1,\n",
    "    )\n",
    "    ax[mask_index].hist(\n",
    "        latent_stds_masked[q_masked][mask_index][0, latent_index_m, t_min:t_max],\n",
    "        bins=bins,\n",
    "        color=\"darkred\",\n",
    "        histtype=\"step\",\n",
    "        lw=1,\n",
    "    )\n",
    "    ax[mask_index].hist(\n",
    "        latent_stds_naive[q_naive][mask_index][0, latent_index_n, t_min:t_max],\n",
    "        bins=bins,\n",
    "        color=blues[mask_index],\n",
    "        alpha=0.5,\n",
    "    )\n",
    "    ax[mask_index].hist(\n",
    "        latent_stds_masked[q_masked][mask_index][0, latent_index_m, t_min:t_max],\n",
    "        bins=bins,\n",
    "        color=reds[mask_index],\n",
    "        alpha=0.5,\n",
    "    )\n",
    "\n",
    "    ax[mask_index].set_title(\n",
    "        test_masks[mask_index].replace(\"_\", \" \").replace(\"xa m\", \"mskd\")\n",
    "    )\n",
    "    ax[mask_index].set_xlabel(\"latent std\")\n",
    "ax[0].set_xticks([0.3, 0.9])\n",
    "ax[0].set_ylabel(\"count\")\n",
    "for axi in ax:\n",
    "    axi.spines[\"left\"].set_position((\"outward\", 4))\n",
    "    axi.spines[\"bottom\"].set_position((\"outward\", 4))\n",
    "\n",
    "plt.savefig(\n",
    "    f\"{local_dir}/figures/latent_std_hist_masked_vs_naive_{q_masked}_{q_naive}.png\",\n",
    "    bbox_inches=\"tight\",\n",
    ")\n",
    "plt.savefig(\n",
    "    f\"{local_dir}/figures/latent_std_hist_masked_vs_naive_{q_masked}_{q_naive}.pdf\",\n",
    "    bbox_inches=\"tight\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train a linear ridge regression model from latents to velocity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define test masks\n",
    "test_masks = [\n",
    "    \"all_obs\",\n",
    "    \"xa_m_5\",\n",
    "    \"xa_m_20\",\n",
    "    \"xa_m_50\",\n",
    "    \"xa_m_100\",\n",
    "    \"xa_m_150\",\n",
    "    \"xa_m_200\",\n",
    "]\n",
    "\n",
    "arr = np.zeros(\n",
    "    (len(run_dirs), len(test_masks), 4)\n",
    ")  # q x mask x 4 (train, val, test, test_shuffle)\n",
    "\n",
    "all_results = {\n",
    "    metric: arr.copy()\n",
    "    for metric in [\n",
    "        \"mean_r2\",\n",
    "        \"v_x_r2\",\n",
    "        \"v_y_r2\",\n",
    "        \"v_x_rmse\",\n",
    "        \"v_y_rmse\",\n",
    "        \"v_x_corr\",\n",
    "        \"v_y_corr\",\n",
    "    ]\n",
    "}\n",
    "all_results_spikes = {\n",
    "    metric: arr.copy()\n",
    "    for metric in [\n",
    "        \"mean_r2\",\n",
    "        \"v_x_r2\",\n",
    "        \"v_y_r2\",\n",
    "        \"v_x_rmse\",\n",
    "        \"v_y_rmse\",\n",
    "        \"v_x_corr\",\n",
    "        \"v_y_corr\",\n",
    "    ]\n",
    "}\n",
    "\n",
    "all_variances = np.zeros((len(run_dirs), len(test_masks), model.net.n_latents))\n",
    "all_means = np.zeros((len(run_dirs), len(test_masks), model.net.n_latents))\n",
    "\n",
    "selected_variances = np.zeros((len(run_dirs), len(test_masks)))\n",
    "avg_latents = 1\n",
    "\n",
    "for q in range(len(run_dirs)):\n",
    "    exp_dir = temp_dir + run_dirs[q] + \"/\"\n",
    "\n",
    "    with open(exp_dir + \"model.pkl\", \"rb\") as f:\n",
    "        model = torch.load(f)\n",
    "\n",
    "    # Decode latents for train, validation, and test sets\n",
    "    train_latent_means, train_latent_stds, train_y, train_d, train_spikes_matrices = (\n",
    "        infer_latents_monkey(PD_train, model, test_masks)\n",
    "    )\n",
    "    valid_latent_means, valid_latent_stds, valid_y, valid_d, valid_spikes_matrices = (\n",
    "        infer_latents_monkey(PD_valid, model, test_masks)\n",
    "    )\n",
    "    test_latent_means, test_latent_stds, test_y, test_d, test_spikes_matrices = (\n",
    "        infer_latents_monkey(PD_test, model, test_masks)\n",
    "    )\n",
    "\n",
    "    # standardize the means and the behavioral data\n",
    "\n",
    "    sorted_indices = np.argsort(\n",
    "        np.std(train_latent_means[test_masks.index(\"all_obs\")][0], axis=-1)\n",
    "    )[::-1]\n",
    "    latent_indices = sorted_indices[:avg_latents]  # example latent index to plot\n",
    "    # example usage\n",
    "    scaled_data = scale_datasets(\n",
    "        train_d,\n",
    "        valid_d,\n",
    "        test_d,\n",
    "        train_latent_means,\n",
    "        valid_latent_means,\n",
    "        test_latent_means,\n",
    "        train_spikes_matrices,\n",
    "        valid_spikes_matrices,\n",
    "        test_spikes_matrices,\n",
    "        test_masks,\n",
    "        standardize=False,\n",
    "        min_max=True,\n",
    "    )\n",
    "\n",
    "    (\n",
    "        train_d,\n",
    "        valid_d,\n",
    "        test_d,\n",
    "        train_latent_means,\n",
    "        valid_latent_means,\n",
    "        test_latent_means,\n",
    "        train_spikes_matrices,\n",
    "        valid_spikes_matrices,\n",
    "        test_spikes_matrices,\n",
    "    ) = scaled_data\n",
    "\n",
    "    # train the model on the all observed training data\n",
    "\n",
    "    reg_x_means, reg_y_means = train_on_all_obs(\n",
    "        train_latent_means[test_masks.index(\"all_obs\")],\n",
    "        train_d,\n",
    "        delay=2,\n",
    "        reg_type=\"ridge\",\n",
    "    )\n",
    "    reg_x_d, reg_y_d = train_on_all_obs(train_d, train_d, delay=2, reg_type=\"ridge\")\n",
    "    reg_x_spikes, reg_y_spikes = train_on_all_obs(\n",
    "        train_spikes_matrices[test_masks.index(\"all_obs\")],\n",
    "        train_d,\n",
    "        delay=2,\n",
    "        reg_type=\"ridge\",\n",
    "    )\n",
    "\n",
    "    for mm, mask in enumerate(test_masks):\n",
    "\n",
    "        all_variances[q, mm, :] = np.mean(train_latent_stds[mm], axis=-1)\n",
    "        all_means[q, mm, :] = np.mean(train_latent_means[mm], axis=-1)\n",
    "        selected_variances[q, mm] = np.mean(train_latent_stds[mm][0, latent_indices, :])\n",
    "        print(np.mean(train_latent_stds[mm][0, latent_indices, :], axis=-1))\n",
    "        # now predict and evaluate the models\n",
    "        # latent means to velocity\n",
    "\n",
    "        results_means_train = predict_and_evaluate(\n",
    "            reg_x_means,\n",
    "            reg_y_means,\n",
    "            train_latent_means[test_masks.index(mask)],\n",
    "            train_d,\n",
    "            delay=2,\n",
    "            print_shapes=True,\n",
    "        )\n",
    "        results_means_val = predict_and_evaluate(\n",
    "            reg_x_means,\n",
    "            reg_y_means,\n",
    "            valid_latent_means[test_masks.index(mask)],\n",
    "            valid_d,\n",
    "            delay=2,\n",
    "            print_shapes=True,\n",
    "        )\n",
    "        results_means_test = predict_and_evaluate(\n",
    "            reg_x_means,\n",
    "            reg_y_means,\n",
    "            test_latent_means[test_masks.index(mask)],\n",
    "            test_d,\n",
    "            delay=2,\n",
    "            print_shapes=True,\n",
    "        )\n",
    "        results_means_test_shuffle = predict_and_evaluate(\n",
    "            reg_x_means,\n",
    "            reg_y_means,\n",
    "            test_latent_means[test_masks.index(mask)],\n",
    "            copy_and_shuffle(test_d),\n",
    "            delay=2,\n",
    "            print_shapes=True,\n",
    "        )\n",
    "\n",
    "        for key in all_results.keys():\n",
    "            all_results[key][q, mm, 0] = results_means_train[key]\n",
    "            all_results[key][q, mm, 1] = results_means_val[key]\n",
    "            all_results[key][q, mm, 2] = results_means_test[key]\n",
    "            all_results[key][q, mm, 3] = results_means_test_shuffle[key]\n",
    "        # spikes to velocity\n",
    "\n",
    "        results_spikes_train = predict_and_evaluate(\n",
    "            reg_x_spikes,\n",
    "            reg_y_spikes,\n",
    "            train_spikes_matrices[test_masks.index(mask)],\n",
    "            train_d,\n",
    "            delay=2,\n",
    "            print_shapes=True,\n",
    "        )\n",
    "        results_spikes_val = predict_and_evaluate(\n",
    "            reg_x_spikes,\n",
    "            reg_y_spikes,\n",
    "            valid_spikes_matrices[test_masks.index(mask)],\n",
    "            valid_d,\n",
    "            delay=2,\n",
    "            print_shapes=True,\n",
    "        )\n",
    "        results_spikes_test = predict_and_evaluate(\n",
    "            reg_x_spikes,\n",
    "            reg_y_spikes,\n",
    "            test_spikes_matrices[test_masks.index(mask)],\n",
    "            test_d,\n",
    "            delay=2,\n",
    "            print_shapes=True,\n",
    "        )\n",
    "        results_spikes_test_shuffle = predict_and_evaluate(\n",
    "            reg_x_spikes,\n",
    "            reg_y_spikes,\n",
    "            test_spikes_matrices[test_masks.index(mask)],\n",
    "            copy_and_shuffle(test_d),\n",
    "            delay=2,\n",
    "            print_shapes=True,\n",
    "        )\n",
    "\n",
    "        for key in all_results_spikes.keys():\n",
    "            all_results_spikes[key][q, mm, 0] = results_spikes_train[key]\n",
    "            all_results_spikes[key][q, mm, 1] = results_spikes_val[key]\n",
    "            all_results_spikes[key][q, mm, 2] = results_spikes_test[key]\n",
    "            all_results_spikes[key][q, mm, 3] = results_spikes_test_shuffle[key]\n",
    "        # reference model to check correctness velocity from velocity\n",
    "\n",
    "    results_d_train = predict_and_evaluate(\n",
    "        reg_x_d, reg_y_d, train_d, train_d, delay=2, print_shapes=True\n",
    "    )\n",
    "    results_d_val = predict_and_evaluate(\n",
    "        reg_x_d, reg_y_d, valid_d, valid_d, delay=2, print_shapes=True\n",
    "    )\n",
    "    results_d_test = predict_and_evaluate(\n",
    "        reg_x_d, reg_y_d, test_d, test_d, delay=2, print_shapes=True\n",
    "    )\n",
    "    results_d_test_shuffle = predict_and_evaluate(\n",
    "        reg_x_d, reg_y_d, copy_and_shuffle(test_d), test_d, delay=2, print_shapes=True\n",
    "    )\n",
    "\n",
    "    print(model_tags[q])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Figure 6 Panel E:\n",
    "### Latent uncertainty vs. decoding performance for many model seeds fit with ridge regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Naive Index 0 - Slope: 0.035, p-value: 0.458\n",
      "Naive Index 1 - Slope: 0.179, p-value: 0.040\n",
      "Naive Index 2 - Slope: 0.222, p-value: 0.002\n",
      "Naive Index 3 - Slope: 0.068, p-value: 0.123\n",
      "Naive Index 4 - Slope: 0.139, p-value: 0.139\n",
      "Naive Index 5 - Slope: 0.035, p-value: 0.008\n",
      "Naive Index 6 - Slope: 0.117, p-value: 0.158\n",
      "Naive Index 7 - Slope: 0.079, p-value: 0.030\n",
      "Naive Index 8 - Slope: 0.056, p-value: 0.217\n",
      "Naive Index 9 - Slope: 0.039, p-value: 0.251\n",
      "Masked Index 0 - Slope: -1.611, p-value: 0.001\n",
      "Masked Index 1 - Slope: -1.718, p-value: 0.000\n",
      "Masked Index 2 - Slope: -2.568, p-value: 0.003\n",
      "Masked Index 3 - Slope: -2.186, p-value: 0.004\n",
      "Masked Index 4 - Slope: -1.763, p-value: 0.004\n",
      "Masked Index 5 - Slope: -2.225, p-value: 0.000\n",
      "Masked Index 6 - Slope: -1.483, p-value: 0.000\n",
      "Masked Index 7 - Slope: -1.801, p-value: 0.001\n",
      "Masked Index 8 - Slope: -2.849, p-value: 0.000\n",
      "Masked Index 9 - Slope: -1.651, p-value: 0.001\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 157.48x157.48 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=cm2inch(4, 4))\n",
    "metric = \"v_y_corr\"\n",
    "\n",
    "greys = plt.get_cmap(\"Greys\")(np.linspace(0.3, 0.95, len(test_masks)))\n",
    "\n",
    "masked_indices = [i for i, x in enumerate(model_tags) if x == \"masked\"]\n",
    "naive_indices = [i for i, x in enumerate(model_tags) if x == \"naive\"]\n",
    "# loop through each naive index\n",
    "for i, idx in enumerate(naive_indices):\n",
    "    metric_naive = all_results[metric][idx, :, 2].flatten()\n",
    "    std_naive = selected_variances[idx, :].flatten()\n",
    "\n",
    "    # calculate slope, intercept, and p-value\n",
    "    slope_naive, intercept_naive, r_value_naive, p_value_naive, std_err_naive = (\n",
    "        stats.linregress(metric_naive, std_naive)\n",
    "    )\n",
    "\n",
    "    # plot data with linear fit\n",
    "    for mm in range(len(test_masks)):\n",
    "        plt.plot(\n",
    "            metric_naive[mm],\n",
    "            std_naive[mm],\n",
    "            \"o\",\n",
    "            ms=3,\n",
    "            color=blues[mm],\n",
    "            alpha=0.8,\n",
    "            label=test_masks[mm] if (i == 0) else None,\n",
    "        )\n",
    "    plt.plot(\n",
    "        metric_naive,\n",
    "        slope_naive * metric_naive + intercept_naive,\n",
    "        color=\"C0\",\n",
    "        alpha=0.4,\n",
    "        lw=0.5,\n",
    "        label=f\"m={slope_naive:.3f} (p={p_value_naive:.3f})\",\n",
    "    )\n",
    "\n",
    "    # print slope and p-value\n",
    "    print(f\"Naive Index {i} - Slope: {slope_naive:.3f}, p-value: {p_value_naive:.3f}\")\n",
    "\n",
    "# loop through each masked index\n",
    "for i, idx in enumerate(masked_indices):\n",
    "    metric_masked = all_results[metric][idx, :, 2].flatten()\n",
    "    std_masked = selected_variances[idx, :].flatten()\n",
    "\n",
    "    # calculate slope, intercept, and p-value\n",
    "    slope_masked, intercept_masked, r_value_masked, p_value_masked, std_err_masked = (\n",
    "        stats.linregress(metric_masked, std_masked)\n",
    "    )\n",
    "\n",
    "    # plot data with linear fit\n",
    "    for mm in range(len(test_masks)):\n",
    "        plt.plot(\n",
    "            metric_masked[mm],\n",
    "            std_masked[mm],\n",
    "            \"o\",\n",
    "            ms=3,\n",
    "            color=reds[mm],\n",
    "            alpha=0.8,\n",
    "            label=test_masks[mm] if (i == 0) else None,\n",
    "        )\n",
    "    plt.plot(\n",
    "        metric_masked,\n",
    "        slope_masked * metric_masked + intercept_masked,\n",
    "        alpha=0.4,\n",
    "        lw=0.5,\n",
    "        color=\"darkred\",\n",
    "        label=f\"m={slope_masked:.3f} (p={p_value_masked:.3f})\",\n",
    "    )\n",
    "\n",
    "    # add labels and legend\n",
    "    plt.xlabel(\"correlation velocity\")\n",
    "    plt.ylabel(\"latent uncertainty (std)\")\n",
    "\n",
    "    # print slope and p-value\n",
    "    print(\n",
    "        f\"Masked Index {i} - Slope: {slope_masked:.3f}, p-value: {p_value_masked:.3f}\"\n",
    "    )\n",
    "\n",
    "# move outward\n",
    "plt.gca().spines[\"left\"].set_position((\"outward\", 4))\n",
    "plt.gca().spines[\"bottom\"].set_position((\"outward\", 4))\n",
    "\n",
    "\n",
    "for mm, mask in enumerate(test_masks):\n",
    "    plt.plot(\n",
    "        all_results_spikes[metric][0, mm, 2],\n",
    "        np.min(selected_variances) - 0.05,\n",
    "        marker=\"|\",\n",
    "        color=greys[mm],\n",
    "        markersize=3,\n",
    "        label=mask.replace(\"_\", \" \").replace(\"xa m\", \"\"),\n",
    "        zorder=0,\n",
    "    )\n",
    "\n",
    "plt.yticks([0.3, 0.7])\n",
    "plt.legend(loc=\"upper right\", bbox_to_anchor=(4, 1), fontsize=6, ncols=3)\n",
    "plt.savefig(\n",
    "    f\"{local_dir}/figures/{metric}_vs_uncertainty_linear_fit_avg_latents_{avg_latents}.png\",\n",
    "    bbox_inches=\"tight\",\n",
    ")\n",
    "plt.savefig(\n",
    "    f\"{local_dir}/figures/{metric}_vs_uncertainty_linear_fit_avg_latents_{avg_latents}.pdf\",\n",
    "    bbox_inches=\"tight\",\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "vaenbc",
   "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": "3.8.19"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}