Figure4.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/lib/python3.7/site-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
      "  import pandas.util.testing as tm\n"
     ]
    }
   ],
   "source": [
    "import math\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "from scipy.stats import nbinom\n",
    "from tqdm.notebook import trange, tqdm\n",
    "import seaborn as sns \n",
    "import pandas as pd\n",
    "import scipy.stats as stats\n",
    "\n",
    "#%matplotlib notebook\n",
    "\n",
    "plt.rcParams.update({\n",
    "    \"text.usetex\": True,\n",
    "    \"font.family\": \"serif\",\n",
    "    \"font.serif\": [\"Helvetica\"]})\n",
    "\n",
    "plt.rc('xtick', labelsize = 25)\n",
    "plt.rc('ytick', labelsize = 25)\n",
    "\n",
    "from matplotlib import cm, colors, colorbar\n",
    "cmap = cm.get_cmap('viridis_r')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to scale-up the colormap regarding the age of the individual\n",
    "def trans(x):\n",
    "    a = 18/57\n",
    "    b = 1\n",
    "    return (x-a)/(b-a)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "All the inference analysis on real data were performed using the class methods available in utils/full_inference_tools. \n",
    "**Remark**\n",
    "$A = \\tau^{-1}$ and $B = \\theta^{-1}$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Results of the inference for S1, S2, MPbeta, MPalpha, YB\n",
    "###P3\n",
    "A_P3 = np.array([-0.31402038, -0.31402038, -0.27625993, -0.2567954 , -0.23052102,\n",
    "       -0.2251288 , -0.23053352, -0.25110638, -0.28754419, -0.3609712 ,\n",
    "       -0.46874922, -0.57265123, -0.60609602, -0.53379117])\n",
    "A_P3_err = np.array([0.03010724, 0.03010724, 0.02885052, 0.02938364, 0.02856642,\n",
    "       0.02866415, 0.03365785, 0.03225998, 0.03837605, 0.05427912,\n",
    "       0.09668126, 0.1135252 , 0.15082407, 0.11357856])\n",
    "B_P3 = np.array([0.20939261, 0.20939261, 0.21690351, 0.22393034, 0.23294158,\n",
    "       0.2402579 , 0.25172744, 0.28241968, 0.3537332 , 0.47468435,\n",
    "       0.66777199, 0.8346176 , 0.8789646 , 0.83831288])\n",
    "B_P3_err = np.array([0.02419785, 0.02419785, 0.02613197, 0.02530497, 0.03086372,\n",
    "       0.03541774, 0.0348638 , 0.04664989, 0.0694717 , 0.10337856,\n",
    "       0.1995851 , 0.16047709, 0.20008823, 0.13579055])\n",
    "\n",
    "###P4\n",
    "A_P4 = np.array([-0.33899975, -0.30272089, -0.26214746, -0.24277183, -0.23834802,\n",
    "       -0.22568074, -0.25099427, -0.32360902, -0.34902087, -0.43101271,\n",
    "       -0.52035608, -0.83375736, -0.83241119, -0.62396399])\n",
    "A_P4_err = np.array([0.0085284 , 0.00791802, 0.00449601, 0.00912208, 0.00515741,\n",
    "       0.01514763, 0.02773582, 0.04871143, 0.01932869, 0.03780793,\n",
    "       0.0277149 , 0.08536917, 0.06369186, 0.25750091])\n",
    "\n",
    "B_P4 = np.array([0.45074493, 0.46225775, 0.46967356, 0.4758764 , 0.49529015,\n",
    "       0.48245649, 0.5418097 , 0.87944082, 0.98517916, 1.0542311 ,\n",
    "       1.21886515, 1.47785786, 1.33087599, 1.22611019])\n",
    "B_P4_err = np.array([0.01145982, 0.00805639, 0.01224395, 0.01953588, 0.00780943,\n",
    "       0.02658801, 0.03595823, 0.17410984, 0.09481136, 0.05435831,\n",
    "       0.21137076, 0.37766387, 0.28862259, 0.18677284])\n",
    "\n",
    "###P6 beta\n",
    "A_P6_beta = np.array([-0.14707124, -0.14707124, -0.11534869, -0.08613379, -0.08255328,\n",
    "       -0.09177318, -0.10297632, -0.10008767, -0.08187944, -0.04767768,\n",
    "       -0.04188368,  0.02018629,  0.02693291, -0.00238157])\n",
    "A_P6_beta_err = np.array([0.04860392, 0.04860392, 0.01076118, 0.00829211, 0.00547916,\n",
    "       0.00378921, 0.00363806, 0.00962537, 0.01078864, 0.01293259,\n",
    "       0.01378962, 0.02050525, 0.00927396, 0.00719821])\n",
    "\n",
    "B_P6_beta = np.array([0.1409258 , 0.1409258 , 0.19709249, 0.19622967, 0.20407467,\n",
    "       0.21164949, 0.22917316, 0.25016793, 0.26955572, 0.25476055,\n",
    "       0.23725685, 0.16225359, 0.11834641, 0.1047153 ])\n",
    "B_P6_beta_err = np.array([0.08279192, 0.08279192, 0.02995125, 0.03153803, 0.03144363,\n",
    "       0.0264483 , 0.02672435, 0.02587421, 0.03005633, 0.04051904,\n",
    "       0.04361805, 0.04208993, 0.02141383, 0.0179099 ])\n",
    "\n",
    "###P6 alpha\n",
    "A_P6_alpha = np.array([-0.1247948 , -0.09905249, -0.08891382, -0.0873424 , -0.08732188,\n",
    "       -0.08513647, -0.08197841, -0.05923807, -0.02638299,  0.01387177,\n",
    "        0.05756348,  0.10339188,  0.1064932 ,  0.10130882])\n",
    "\n",
    "A_P6_alpha_err = np.array([0.00689288, 0.01060347, 0.01087509, 0.01169527, 0.0081088 ,\n",
    "       0.01037073, 0.0119971 , 0.0126808 , 0.01374629, 0.01224552,\n",
    "       0.01711305, 0.01562535, 0.01152943, 0.00875642])\n",
    "\n",
    "B_P6_alpha = np.array([0.16655104, 0.16866091, 0.173838  , 0.18420626, 0.19626572,\n",
    "       0.21766389, 0.23922646, 0.24558271, 0.23467734, 0.1908341 ,\n",
    "       0.15612782, 0.12121458, 0.10963785, 0.09227638])\n",
    "B_P6_alpha_err = np.array([0.01702224, 0.01976717, 0.01949749, 0.02086341, 0.02224078,\n",
    "       0.02538924, 0.03119845, 0.03608516, 0.04459864, 0.0324551 ,\n",
    "       0.03078455, 0.03266058, 0.02095319, 0.01003768])\n",
    "\n",
    "###P9\n",
    "A_P9 = np.array([-0.11728954, -0.04835404, -0.01823364, -0.02163358, -0.0207305 ,\n",
    "       -0.02923776, -0.03938855, -0.05070962, -0.05013025, -0.06651899,\n",
    "       -0.06613541, -0.03882864, -0.06743096, -0.12951165])\n",
    "A_P9_err = np.array([0.00628255, 0.01089386, 0.01935156, 0.01316055, 0.01489642,\n",
    "       0.01417742, 0.0162307 , 0.01861891, 0.02180234, 0.02442195,\n",
    "       0.03201769, 0.03449101, 0.02543406, 0.01881537])\n",
    "\n",
    "B_P9 = np.array([0.09211787, 0.08645331, 0.03690246, 0.08515661, 0.08980447,\n",
    "       0.09881692, 0.10447173, 0.10567664, 0.1181512 , 0.13879699,\n",
    "       0.14025146, 0.10645719, 0.0917193 , 0.06787061])\n",
    "B_P9_err = np.array([0.01906303, 0.01846389, 0.02911413, 0.01923114, 0.02054743,\n",
    "       0.02188888, 0.02428085, 0.02374876, 0.02523934, 0.02800367,\n",
    "       0.02386697, 0.02077303, 0.01609406, 0.00976362])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#fthresh = 1E-5 \n",
    "A_thresh_1em5 = np.array([ A_P3[5], A_P4[5], A_P6_beta[4], A_P6_alpha[3], A_P9[4]])\n",
    "B_thresh_1em5 = np.array([ B_P3[5], B_P4[5], B_P6_beta[4], B_P6_alpha[3], B_P9[4]])\n",
    "\n",
    "A_thresh_err_1em5 = np.array([A_P3_err[5], A_P4_err[5], A_P6_beta_err[4], A_P6_alpha_err[3], A_P9_err[4]])\n",
    "B_thresh_err_1em5 = np.array([B_P3_err[5], B_P4_err[5], B_P6_beta_err[4], B_P6_alpha_err[3], B_P9_err[4]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "f_thresh_vec_twins = np.logspace(-6, -3.5, 14)\n",
    "f_thresh_vec_MP = np.logspace(-5.8, -3.5, 14)\n",
    "f_thresh_vec_MP_alpha = np.logspace(-5.4, -3.5, 14)\n",
    "f_thresh_vec_YB = np.logspace(-5.6, -3.5, 14)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build the dataframe without HR\n",
    "A = -A_thresh_1em5\n",
    "B = B_thresh_1em5 \n",
    "\n",
    "A_err = A_thresh_err_1em5\n",
    "B_err = B_thresh_err_1em5\n",
    "\n",
    "d = {'A': A, 'B': B , 'A_err' : A_err, 'B_err' : B_err}\n",
    "df_fthresh_1m5 = pd.DataFrame(data = d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#add value of the colormap index\n",
    "df_fthresh_1m5['cmap'] = np.array([ trans(21/57), trans(21/57), trans(28/57), trans(28/57), trans(57/57)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## B/ alpha_theo vs alpha_inf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def alpha_err_func(A, dA, B, dB):\n",
    "    return (2/B)*dA + np.abs(2*A/(B**2))*dB\n",
    "A_thresh_1em5, B_thresh_1em5 = df_fthresh_1m5['A'], df_fthresh_1m5['B']\n",
    "A_thresh_err_1em5, B_hresh_err_1em5 = df_fthresh_1m5['A_err'], df_fthresh_1m5['B_err'] \n",
    "alpha_dyn_inf = 2*A_thresh_1em5/B_thresh_1em5\n",
    "alpha_dyn_err = alpha_err_func(A_thresh_1em5, A_thresh_err_1em5, B_thresh_1em5, B_thresh_err_1em5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "alpha_null_P9 = np.mean([-2.12554061, -2.02543194])\n",
    "#alpha_null_Patient_3 = np.mean([np.array([-0.9924140178297096, -1.0392001407555118, -1.0008845465713057, -1.0080247654349548,-1.0173379589080598, -1.0027652402045957, -1.0157189718977282, -0.9993741280628488])])-1\n",
    "alpha_null_P6_alpha = np.mean([-1.9282954, -1.90240558])\n",
    "alpha_null_P6_beta = np.mean([-2.02192528, -2.01043465])\n",
    "alpha_null_P3 = np.mean([-2.13538507, -2.15983246])\n",
    "alpha_null_P4 = np.mean([-2.196335, -2.12211096, -2.18253516])\n",
    "#alpha_null_Patient_1 = np.mean([np.array([-1.5413041433045693, -1.5189818323965012, -1.5057620821570976, -1.6120918510113957, -1.4364969283797406, -1.4888260781087121, -1.5363600571227614, -1.5035519661480177])])-1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "alpha_null_P9_std = np.std([-2.12554061, -2.02543194])\n",
    "#alpha_null_Patient_3_std = np.std([np.array([-0.9924140178297096, -1.0392001407555118, -1.0008845465713057, -1.0080247654349548,-1.0173379589080598, -1.0027652402045957, -1.0157189718977282, -0.9993741280628488])])\n",
    "alpha_null_P6_alpha_std = np.std([-1.9282954, -1.90240558])\n",
    "alpha_null_P6_beta_std = np.std([-2.02192528, -2.01043465])\n",
    "alpha_null_P3_std = np.std([-2.13538507, -2.15983246])\n",
    "alpha_null_P4_std = np.std([-2.196335, -2.12211096, -2.18253516])\n",
    "#alpha_null_Patient_1_std = np.std([np.array([-1.5413041433045693, -1.5189818323965012, -1.5057620821570976, -1.6120918510113957, -1.4364969283797406, -1.4888260781087121, -1.5363600571227614, -1.5035519661480177])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "alpha_null = np.array([alpha_null_P9, alpha_null_P6_alpha, alpha_null_P6_beta, alpha_null_P4, alpha_null_P3])\n",
    "alpha_null_err = np.array([alpha_null_P9_std, alpha_null_P6_alpha_std, alpha_null_P6_beta_std, alpha_null_P4_std, alpha_null_P3_std])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "alpha_null = np.flip(alpha_null)\n",
    "alpha_null_err = np.flip(alpha_null_err)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_fthresh_1m5['alpha_dyn_inf'] = alpha_dyn_inf\n",
    "df_fthresh_1m5['alpha_dyn_err'] = alpha_dyn_err\n",
    "df_fthresh_1m5['alpha_null'] = -(alpha_null+1)\n",
    "df_fthresh_1m5['alpha_null_err'] = -(alpha_null_err+1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## C/ tau vs age"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "tau = 1/A_thresh_1em5\n",
    "tau_err = np.array(A_thresh_err_1em5)/(np.array(A_thresh_1em5)**2)\n",
    "\n",
    "#age\n",
    "age = np.array([21, 21, 28, 28, 57])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Persistence Probability"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Theory predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "#P3 theory predictions\n",
    "P_pers_P3_F1_F1 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S1_F1_F1.npy')\n",
    "P_pers_P3_F1_F2 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S1_F1_F2.npy')\n",
    "P_pers_P3_F2_F1 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S1_F2_F1.npy')\n",
    "P_pers_P3_F2_F2 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S1_F2_F2.npy')\n",
    "\n",
    "P_pers_P3_F1_F1_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S1_F1_F1_minus.npy')\n",
    "P_pers_P3_F1_F2_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S1_F1_F2_minus.npy')\n",
    "P_pers_P3_F2_F1_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S1_F2_F1_minus.npy')\n",
    "P_pers_P3_F2_F2_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S1_F2_F2_minus.npy')\n",
    "\n",
    "P_pers_P3_F1_F1_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S1_F1_F1_plus.npy')\n",
    "P_pers_P3_F1_F2_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S1_F1_F2_plus.npy')\n",
    "P_pers_P3_F2_F1_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S1_F2_F1_plus.npy')\n",
    "P_pers_P3_F2_F2_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S1_F2_F2_plus.npy')\n",
    "\n",
    "#P4 theory predictions\n",
    "P_pers_P4_F1_F1 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S2_F1_F1.npy')\n",
    "P_pers_P4_F1_F2 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S2_F1_F2.npy')\n",
    "P_pers_P4_F2_F1 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S2_F2_F1.npy')\n",
    "P_pers_P4_F2_F2 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S2_F2_F2.npy')\n",
    "\n",
    "P_pers_P4_F1_F1_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S2_F1_F1_minus.npy')\n",
    "P_pers_P4_F1_F2_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S2_F1_F2_minus.npy')\n",
    "P_pers_P4_F2_F1_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S2_F2_F1_minus.npy')\n",
    "P_pers_P4_F2_F2_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S2_F2_F2_minus.npy')\n",
    "\n",
    "P_pers_P4_F1_F1_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S2_F1_F1_plus.npy')\n",
    "P_pers_P4_F1_F2_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S2_F1_F2_plus.npy')\n",
    "P_pers_P4_F2_F1_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S2_F2_F1_plus.npy')\n",
    "P_pers_P4_F2_F2_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_S2_F2_F2_plus.npy')\n",
    "\n",
    "#P6beta \n",
    "\n",
    "P_pers_P6b_F1_F1 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_MPb_F1_F1.npy')\n",
    "P_pers_P6b_F1_F2 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_MPb_F1_F2.npy')\n",
    "P_pers_P6b_F2_F1 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_MPb_F2_F1.npy')\n",
    "P_pers_P6b_F2_F2 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_MPb_F2_F2.npy')\n",
    "\n",
    "P_pers_P6b_F1_F1_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_MPb_F1_F1_minus.npy')\n",
    "P_pers_P6b_F1_F2_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_MPb_F1_F2_minus.npy')\n",
    "P_pers_P6b_F2_F1_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_MPb_F2_F1_minus.npy')\n",
    "P_pers_P6b_F2_F2_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_MPb_F2_F2_minus.npy')\n",
    "\n",
    "P_pers_P6b_F1_F1_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_MPb_F1_F1_plus.npy')\n",
    "P_pers_P6b_F1_F2_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_MPb_F1_F2_plus.npy')\n",
    "P_pers_P6b_F2_F1_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_MPb_F2_F1_plus.npy')\n",
    "P_pers_P6b_F2_F2_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_MPb_F2_F2_plus.npy')\n",
    "\n",
    "\n",
    "#P9\n",
    "P_pers_P9_bis_F1 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_YB_bis_F1.npy')\n",
    "P_pers_P9_bis_F2 = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_YB_bis_F2.npy')\n",
    "\n",
    "P_pers_P9_F1_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_YBb_F1_F1_minus.npy')\n",
    "P_pers_P9_F2_minus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_YBb_F1_F2_minus.npy')\n",
    "\n",
    "P_pers_P9_F1_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_YBb_F1_F1_plus.npy')\n",
    "P_pers_P9_F2_plus = np.load('/Users/bensouda/Documents/PhD/TCR-dynamics/Codes_and_notebooks/P_pers/P_pers_YBb_F1_F2_plus.npy')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "P_pers_P3 = np.vstack([P_pers_P3_F1_F1, P_pers_P3_F1_F2, P_pers_P3_F2_F1, P_pers_P3_F2_F2])\n",
    "P_pers_P4 = np.vstack([P_pers_P4_F1_F1, P_pers_P4_F1_F2, P_pers_P4_F2_F1, P_pers_P4_F2_F2])\n",
    "P_pers_P6b = np.vstack([P_pers_P6b_F1_F1, P_pers_P6b_F1_F2])\n",
    "P_pers_P9 = np.vstack([P_pers_P9_F1, P_pers_P9_F2])\n",
    "\n",
    "P_pers_P3_plus = np.vstack([P_pers_P3_F1_F1_plus, P_pers_P3_F1_F2_plus, P_pers_P3_F2_F1_plus, P_pers_P3_F2_F2_plus])\n",
    "P_pers_P4_plus = np.vstack([P_pers_P4_F1_F1_plus, P_pers_P4_F1_F2_plus, P_pers_P4_F2_F1_plus, P_pers_P4_F2_F2_plus])\n",
    "P_pers_P6b_plus = np.vstack([P_pers_P6b_F1_F1_plus, P_pers_P6b_F1_F2_plus, P_pers_P6b_F2_F1_plus, P_pers_P6b_F2_F2_plus])\n",
    "P_pers_P9_plus = np.vstack([P_pers_P9_F1_plus, P_pers_P9_F2_plus])\n",
    "\n",
    "P_pers_P3_minus = np.vstack([P_pers_P3_F1_F1_minus, P_pers_P3_F1_F2_minus, P_pers_P3_F2_F1_minus, P_pers_P3_F2_F2_minus])\n",
    "P_pers_P4_minus = np.vstack([P_pers_P4_F1_F1_minus, P_pers_P4_F1_F2_minus, P_pers_P4_F2_F1_minus, P_pers_P4_F2_F2_minus])\n",
    "P_pers_P6b_minus = np.vstack([P_pers_P6b_F1_F1_minus, P_pers_P6b_F1_F2_minus, P_pers_P6b_F2_F1_minus, P_pers_P6b_F2_F2_minus])\n",
    "P_pers_P9_minus = np.vstack([P_pers_P9_F1_minus, P_pers_P9_F2_minus])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "P_pers_P3 = np.vstack([P_pers_P3_F1_F1, P_pers_P3_F1_F2, P_pers_P3_F2_F1, P_pers_P3_F2_F2])\n",
    "P_pers_P3_mean = np.mean(P_pers_P3, axis = 0)\n",
    "P_pers_P3_plus = np.vstack([P_pers_P3_F1_F1_plus, P_pers_P3_F1_F2_plus, P_pers_P3_F2_F1_plus, P_pers_P3_F2_F2_plus])\n",
    "P_pers_P3_minus = np.vstack([P_pers_P3_F1_F1_minus, P_pers_P3_F1_F2_minus, P_pers_P3_F2_F1_minus, P_pers_P3_F2_F2_minus])\n",
    "P_pers_P3_std_plus = np.mean(P_pers_P3_plus, axis = 0)\n",
    "P_pers_P3_std_minus = np.mean(P_pers_P3_minus, axis = 0)\n",
    "\n",
    "P_pers_P3_mean = np.mean(P_pers_P3, axis = 0)\n",
    "P_pers_P4_mean = np.mean(P_pers_P4, axis = 0)\n",
    "P_pers_P6b_mean = np.mean(P_pers_P6b, axis = 0)\n",
    "P_pers_P9_mean = np.mean(P_pers_P9, axis = 0)\n",
    "\n",
    "P_pers_P3_std_plus = np.mean(P_pers_P3_plus, axis = 0)\n",
    "P_pers_P4_std_plus = np.mean(P_pers_P4_plus, axis = 0)\n",
    "P_pers_P6b_std_plus = np.mean(P_pers_MP6b_plus, axis = 0)\n",
    "P_pers_P9_std_plus = np.mean(P_pers_P9_plus, axis = 0)\n",
    "\n",
    "P_pers_P3_std_minus = np.mean(P_pers_P3_minus, axis = 0)\n",
    "P_pers_P4_std_minus = np.mean(P_pers_P4_minus, axis = 0)\n",
    "P_pers_P6b_std_minus = np.mean(P_pers_P6b_minus, axis = 0)\n",
    "P_pers_P9_std_minus = np.mean(P_pers_P9_minus, axis = 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Empirical predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "Pers_emp_P3_F1_0, Pers_emp_P3 = 0.9939736346516007, 0.6995876278695606\n",
    "Pers_emp_P4_F1_0, Pers_emp_P4 = 0.99721059972106, 0.7884650725438299\n",
    "Pers_emp_P6_0, Pers_emp_P6beta = 0.9987974462518869, 0.9844078156132762\n",
    "Pers_emp_P9_F1_0, Pers_emp_P9 = 0.9991653915704548, 0.9791740622411549"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Complete figure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.\n",
      "  x[:, None]\n",
      "/opt/anaconda3/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.\n",
      "  y = y[:, np.newaxis]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x1080 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(2,2,figsize=(18,15))\n",
    "plt.rc('text', usetex=True)\n",
    "plt.rcParams.update({\n",
    "    \"text.usetex\": True,\n",
    "    \"font.family\": \"serif\",\n",
    "    \"font.serif\": [\"Helvetica\"]})\n",
    "\n",
    "\n",
    "# Subplot 1\n",
    "\n",
    "X = np.linspace(0, 0.3, 100)\n",
    "Y = 2*X\n",
    "\n",
    "# New groups without HR\n",
    "groups = {r'P05 $\\beta$, 20-23/25 years old': (-0.35,0.72),\n",
    "          r'P04 $\\beta$, 20-23/25 years old': (0,0.08),\n",
    "           r'P06 $\\beta$, 31-32 years old' : (-0.1, 0.38),\n",
    "            r'P06 $\\alpha$, 31-32 years old': (-0.2, 0.25),\n",
    "            r'P03 $\\beta$, 57-58 years old': (-0.05, 0.2)}\n",
    "\n",
    "axs[0,0].plot(X,Y, linestyle = 'dotted', color = 'DarkGrey', label =  r'$ 1/ \\theta=2/ \\tau$') \n",
    "for i, label in enumerate(groups.keys()):\n",
    "    axs[0,0].errorbar(df_fthresh_1m5.loc[df_fthresh_1m5['label']==label, 'A'], df_fthresh_1m5.loc[df_fthresh_1m5['label']==label,'B'], xerr = df_fthresh_1m5.loc[df_fthresh_1m5['label']==label, 'A_err'], yerr= df_fthresh_1m5.loc[df_fthresh_1m5['label']==label, 'B_err'] , color = cmap(float(df_fthresh_1m5.loc[df_fthresh_1m5['label']==label, 'cmap'])), capsize = 10, linewidth = 3)\n",
    "\n",
    "axs[0,0].set_xlabel(r'$  \\tau^{-1} \\ (\\rm{year}^{-1})$', fontsize = 35)\n",
    "axs[0,0].set_ylabel(r'$  \\theta^{-1} \\ (\\rm{year}^{-1})$', fontsize = 35)\n",
    "axs[0,0].axis([0, 0.3, 0, 0.6])\n",
    "axs[0,0].xaxis.set_tick_params(width=.6)\n",
    "axs[0,0].yaxis.set_tick_params(width=.6)\n",
    "plt.setp(axs[0,0].spines.values(), linewidth=.6)\n",
    "\n",
    "#Subplot 2\n",
    "\n",
    "x = np.linspace(0.8, 1.3)\n",
    "axs[0,1].plot(x,x, linestyle = 'dotted', color = 'DarkGrey') \n",
    "\n",
    "axs[0,1].errorbar(-(alpha_null[0]+1), alpha_dyn_inf[0] , xerr = alpha_null_err[0], yerr = alpha_dyn_err[0] , capsize = 10, color= cmap(trans(21/57)), linestyle= '', marker = '.', markersize = 10, linewidth = 3)\n",
    "axs[0,1].errorbar(-(alpha_null[1]+1), alpha_dyn_inf[1] , xerr = alpha_null_err[1], yerr = alpha_dyn_err[1] , capsize = 10, color= cmap(trans(21/57)), linestyle= '', marker = '.', markersize = 10, linewidth = 3)\n",
    "axs[0,1].errorbar(-(alpha_null[2]+1), alpha_dyn_inf[2] , xerr = alpha_null_err[2], yerr = alpha_dyn_err[2] , capsize = 10, color= cmap(trans(28/57)), linestyle= '', marker = '.', markersize = 10, linewidth = 3)\n",
    "axs[0,1].errorbar(-(alpha_null[3]+1), alpha_dyn_inf[3] , xerr = alpha_null_err[3], yerr = alpha_dyn_err[3] , capsize = 10, color= cmap(trans(28/57)), linestyle= '', marker = '.', markersize = 10, linewidth = 3)\n",
    "axs[0,1].errorbar(-(alpha_null[4]+1), alpha_dyn_inf[4] , xerr = alpha_null_err[4], yerr = alpha_dyn_err[4] , capsize = 10, color= cmap(trans(57/57)), linestyle= '', marker = '.', markersize = 10, linewidth = 3)\n",
    "\n",
    "axs[0,1].set_xlabel(r'$\\alpha$', fontsize = 35)\n",
    "axs[0,1].set_ylabel(r'$ 2\\theta / \\tau$', fontsize = 35)\n",
    "axs[0,1].xaxis.set_tick_params(width=.6)\n",
    "axs[0,1].yaxis.set_tick_params(width=.6)\n",
    "plt.setp(axs[0,1].spines.values(), linewidth=.6)\n",
    "\n",
    "# Subplot 3\n",
    "\n",
    "axs[1,0].plot(age, tau, linestyle= 'dotted', color = 'DarkGrey')\n",
    "axs[1,0].errorbar(age[0], tau[0], yerr = tau_err[0], capsize = 10 , color= cmap(trans(21/57)),  marker =\"1\", markersize = 10, linewidth = 3)\n",
    "axs[1,0].errorbar(age[1], tau[1], yerr = tau_err[1], capsize = 10 , color= cmap(trans(21/57)),  marker =\"1\", markersize = 10, linewidth = 3)\n",
    "axs[1,0].errorbar(age[2], tau[2], yerr = tau_err[2], capsize = 10 ,  color= cmap(trans(28/57)),  marker =\"1\", markersize = 10, linewidth = 3)\n",
    "axs[1,0].errorbar(age[3], tau[3], yerr = tau_err[3], capsize = 10 , color= cmap(trans(28/57)),  marker =\"1\", markersize = 10, linewidth = 3)\n",
    "axs[1,0].errorbar(age[4], tau[4], yerr = tau_err[4], capsize = 10 ,  color= cmap(trans(57/57)),  marker =\"1\", markersize = 10, linewidth = 3)\n",
    "axs[1,0].set_xlabel(r'age (year)', fontsize = 35)\n",
    "axs[1,0].set_ylabel(r'$  \\tau\\ (\\rm{year})$', fontsize = 35)\n",
    "axs[1,0].xaxis.set_tick_params(width=.6)\n",
    "axs[1,0].yaxis.set_tick_params(width=.6)\n",
    "plt.setp(axs[1,0].spines.values(), linewidth=.6)\n",
    "\n",
    "# Subplot 4\n",
    "\n",
    "t_vec = np.linspace(0.1, 10, 10)\n",
    "\n",
    "P3_1 = axs[1,1].errorbar(t_vec, P_pers_P3_mean, yerr = 0 , uplims=False, lolims=False, color = cmap(trans(21/57)), capsize = 1, linestyle= '-', marker = '')\n",
    "P3_2 = axs[1,1].fill_between(t_vec, P_pers_P3_std_minus, P_pers_P3_std_plus  , color = cmap(trans(21/57)), alpha = 0.1)\n",
    "P3_data = axs[1,1].plot([(730-45)/365], [Pers_emp_P3], color = cmap(trans(21/57)), marker = '^', linestyle = ' ',  markersize = 15, alpha = 1, label = r'data')\n",
    "\n",
    "P4_1 = axs[1,1].errorbar(t_vec, P_pers_P4_mean, yerr = 0 , uplims=False, lolims=False, color = cmap(trans(21/57)), capsize = 1, linestyle= '--', marker = '')\n",
    "P4_2 = axs[1,1].fill_between(t_vec, P_pers_P4_std_minus, P_pers_P4_std_plus  , color = cmap(trans(21/57)), alpha = 0.1)\n",
    "P4_data = axs[1,1].plot([(730-45)/365], [Pers_emp_P4], color = cmap(trans(21/57)), marker = '^', linestyle = ' ', markersize = 15, alpha = 1)\n",
    "\n",
    "\n",
    "P6b_1 = axs[1,1].errorbar(t_vec, P_pers_P6b_mean, yerr = 0 , uplims=False, lolims=False, color = cmap(trans(28/57)), capsize = 1, linestyle= '-', marker = '')\n",
    "P6b_2 = axs[1,1].fill_between(t_vec, P_pers_P6b_std_minus, P_pers_P6b_std_plus,  color = cmap(trans(28/57)), alpha =0.1)\n",
    "P6b_data = axs[1,1].plot([1.5], [Pers_emp_P6beta], color = cmap(trans(28/57)), marker = '^', linestyle = ' ', markersize = 15, alpha = 1)\n",
    "\n",
    "\n",
    "P9_1 = axs[1,1].errorbar(t_vec, P_pers_P9_mean, yerr = 0 , uplims=False, lolims=False, color = cmap(trans(57/57)), capsize = 1, linestyle= '-', marker = '')\n",
    "P9_2 = axs[1,1].fill_between(t_vec, P_pers_P9_std_minus, P_pers_P9_std_plus, color= cmap(trans(57/57)), alpha = 0.1)\n",
    "P9_data = axs[1,1].plot([1], [Pers_emp_P9], color = cmap(trans(57/57)), marker = '^', linestyle = ' ', markersize = 15, alpha = 1)\n",
    "\n",
    "axs[1,1].set_xlabel(r'time (year)', fontsize = 35)\n",
    "axs[1,1].set_ylabel(r'persistence probability', fontsize = 35)\n",
    "plt.axis([None,None, 0, 1.1])\n",
    "\n",
    "axs[1,1].yaxis.set_tick_params(width=.8)\n",
    "axs[1,1].legend([(P3_1, P3_2), (P4_1, P4_2), (P6b_1, P6b_2), (P9_1, P9_2)], [r'P3-21 yo', r'P4-21 yo',  r'P6-28 yo', r'P9-57 yo'], fontsize = 20, frameon = False)\n",
    "\n",
    "axs[1,1].xaxis.set_tick_params(width=.6)\n",
    "axs[1,1].yaxis.set_tick_params(width=.6)\n",
    "plt.setp(axs[1,1].spines.values(), linewidth=.6)\n",
    "\n",
    "# set the spacing between subplots\n",
    "plt.subplots_adjust(left=0.1,\n",
    "                    bottom=0.1, \n",
    "                    right=0.9, \n",
    "                    top=0.9, \n",
    "                    wspace=0.3, \n",
    "                    hspace=0.3)\n",
    "\n",
    "fig.savefig('/Users/bensouda/Documents/PhD/Codes_figures/Fig4.pdf', bbox_inches='tight')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}