{"cells": [{"cell_type": "raw", "id": "3ef6e4ee", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Quantifying Accuracy/Precision of Model Parameter Values\n", "########################################################\n", "\n", ".. contents::\n", "\n", "\n", "\n", ".. note::\n", "\n", " Download the example file here: `HP3_TE300_SPC630.hdf5 `_\n", "\n", "First, let's run the following code to generate a basic analysis for us to begin working with.\n", "This code is essential the same as that found in the :doc:`tutorial ` ."]}, {"cell_type": "code", "execution_count": 1, "id": "bec4fd47", "metadata": {}, "outputs": [{"name": "stdout", "output_type": "stream", "text": [" - Optimized (cython) burst search loaded.\n", " - Optimized (cython) photon counting loaded.\n", "--------------------------------------------------------------\n", " You are running FRETBursts (version 0.7.1).\n", "\n", " If you use this software please cite the following paper:\n", "\n", " FRETBursts: An Open Source Toolkit for Analysis of Freely-Diffusing Single-Molecule FRET\n", " Ingargiola et al. (2016). http://dx.doi.org/10.1371/journal.pone.0160716 \n", "\n", "--------------------------------------------------------------\n", "# Total photons (after ALEX selection): 11,414,157\n", "# D photons in D+A excitation periods: 5,208,392\n", "# A photons in D+A excitation periods: 6,205,765\n", "# D+A photons in D excitation period: 6,611,308\n", "# D+A photons in A excitation period: 4,802,849\n", "\n", " - Calculating BG rates ... get bg th arrays\n", "Channel 0\n", "[DONE]\n", " - Performing burst search (verbose=False) ...[DONE]\n", " - Calculating burst periods ...[DONE]\n", " - Counting D and A ph and calculating FRET ... \n", " - Applying background correction.\n", " [DONE Counting D/A]\n"]}, {"data": {"text/plain": ["The model converged after 1 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 36 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 128 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 414 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["2"]}, "execution_count": 1, "metadata": {}, "output_type": "execute_result"}], "source": ["import numpy as np\n", "from matplotlib import pyplot as plt\n", "\n", "import fretbursts as frb\n", "import burstH2MM as bhm\n", "\n", "filename = 'HP3_TE300_SPC630.hdf5'\n", "# load the data into the data object frbdata\n", "frbdata = frb.loader.photon_hdf5(filename)\n", "# if the alternation period is correct, apply data\n", "# plot the alternation histogram\n", "# frb.bpl.plot_alternation_hist(frbdata) # commented so not displayed in notebook\n", "frb.loader.alex_apply_period(frbdata)\n", "# calcualte the background rate\n", "frbdata.calc_bg(frb.bg.exp_fit, F_bg=1.7)\n", "# plot bg parameters, to verify quality\n", "# frb.dplot(frbdata, frb.hist_bg) # commented so not displayed in notebook\n", "# now perform burst search\n", "frbdata.burst_search(m=10, F=6)\n", "# make sure to set the appropriate thresholds of ALL size\n", "# parameters to the particulars of your experiment\n", "frbdata_sel = frbdata.select_bursts(frb.select_bursts.size, th1=50)\n", "# make BurstData object to get data into bursth2MM\n", "bdata = bhm.BurstData(frbdata_sel)\n", "# calculate models\n", "bdata.models.calc_models()"]}, {"cell_type": "raw", "id": "710bb05b", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": [".. _bootstraperr:\n", "\n", "\n", "Bootstrap method\n", "****************\n", "\n", "\n", "\n", "Perhaps the easiest to understand method for quantifying uncertainty in a model is the bootstrap method.\n", "In this method, the bursts are split up into :math:`N`\\ subsets, and separate optimizations are run on each subset.\n", "Then the variance of each parameter value across the :math:`N`\\ different subsets serves as a quantification of the uncertainty.\n", "\n", "In burstH2MM, the :class:`H2MM_result ` object has the :meth:`H2MM_result.bootstrap_eval() ` method which performs this operation.\n", "\n", ":meth:`H2MM_result.bootstrap_eval() ` has one keyword argument: ``subsets`` by which you can specify the number of subsets to divide the data into. \n", "The default is 10, which is usually a good compromise."]}, {"cell_type": "code", "execution_count": 2, "id": "830fecbb", "metadata": {"scrolled": true}, "outputs": [{"data": {"text/plain": ["The model converged after 290 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 226 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 2017 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 227 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 268 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["(array([[89.31520909, 60.39327263, 51.39814368],\n", " [32.52744943, 36.89787477, 24.65788966],\n", " [ 8.20605449, 16.43892423, 12.39367234]]),\n", " array([0.0255139 , 0.00435775, 0.00127755]),\n", " array([0.00960336, 0.00264025, 0.00212474]))"]}, "execution_count": 2, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].bootstrap_eval(subsets=5)"]}, {"cell_type": "raw", "id": "f28ffda6", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Note how the number of subsets is the number of optimizations. This method automatically stores the results in the :attr:`H2MM_result.bootstrap_err ` attribute, and the transition rate, E, and S values are also supplied as return values.\n", "\n", "Once you run bootstrap_eval, you can now access ``trans_std_bs`` , ``E_std_bs`` , ``S_std_bs`` attributes of :class:`H2MM_result ` , which are the standard deviations of each parameter of the optimized models of the subsets."]}, {"cell_type": "code", "execution_count": 3, "id": "344e83ad", "metadata": {}, "outputs": [{"data": {"text/plain": ["array([[89.31520909, 60.39327263, 51.39814368],\n", " [32.52744943, 36.89787477, 24.65788966],\n", " [ 8.20605449, 16.43892423, 12.39367234]])"]}, "execution_count": 3, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].trans_std_bs"]}, {"cell_type": "code", "execution_count": 4, "id": "8de8af42", "metadata": {}, "outputs": [{"data": {"text/plain": ["(array([0.0255139 , 0.00435775, 0.00127755]),\n", " array([0.00960336, 0.00264025, 0.00212474]))"]}, "execution_count": 4, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].E_std_bs, bdata.models[2].S_std_bs"]}, {"cell_type": "raw", "id": "fa3f8fe7", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Closer examination of models\n", "============================\n", "\n", "\n", "\n", ":attr:`H2MM_result.bootstrap_err ` attribute allows closer examination of these subsets.\n", "This attribute is an instance of the class :class:`ModelError.Bootstrap_Error ` , which is made to coordinate the bootstrap evaluation.\n", "The :class:`H2MM_result ` attributes :attr:`H2MM_result.trans_std_bs ` , :attr:`H2MM_result.E_std_bs ` , :attr:`H2MM_result.S_std_bs ` are just aliases of the attributes of :attr:`ModelError.Bootstrap_Error.trans_std ` , :attr:`ModelError.Bootstrap_Error.E_std ` , and :attr:`ModelError.Bootstrap_Error.S_std ` in its :attr:`H2MM_result.bootstrap_err ` attribute."]}, {"cell_type": "code", "execution_count": 5, "id": "b1990f6c", "metadata": {}, "outputs": [{"data": {"text/plain": ["array([[89.31520909, 60.39327263, 51.39814368],\n", " [32.52744943, 36.89787477, 24.65788966],\n", " [ 8.20605449, 16.43892423, 12.39367234]])"]}, "execution_count": 5, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].bootstrap_err.trans_std"]}, {"cell_type": "code", "execution_count": 6, "id": "0a31caff", "metadata": {}, "outputs": [{"data": {"text/plain": ["(array([0.0255139 , 0.00435775, 0.00127755]),\n", " array([0.00960336, 0.00264025, 0.00212474]))"]}, "execution_count": 6, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].bootstrap_err.E_std, bdata.models[2].bootstrap_err.S_std"]}, {"cell_type": "raw", "id": "05af3cfb", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["For E and S, the leakage/direct excitation/ :math:`\\gamma`\\ and :math:`\\beta`\\ correct values: ``ModelError.Boostrap_Error.E_std_corr`` , and ``ModelError.Boostrap_Error.S_std_corr`` "]}, {"cell_type": "code", "execution_count": 7, "id": "20ffd216", "metadata": {}, "outputs": [{"data": {"text/plain": ["(array([0.0255139 , 0.00435775, 0.00127755]),\n", " array([0.00960336, 0.00264025, 0.00212474]))"]}, "execution_count": 7, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].bootstrap_err.E_std_corr, bdata.models[2].bootstrap_err.S_std_corr"]}, {"cell_type": "raw", "id": "344ed7ba", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["If on the other hand, you would prefer to take the standard error, instead of standard deviation of the subsets, there are equivalent attributes :attr:`ModelError.Bootstrap_Error.trans_err ` , :attr:`ModelError.Bootstrap_Error.E_err ` , and :attr:`ModelError.Bootstrap_Error.S_err ` . "]}, {"cell_type": "code", "execution_count": 8, "id": "a54315fa", "metadata": {}, "outputs": [{"data": {"text/plain": ["array([[39.94297579, 27.0086926 , 22.98594864],\n", " [14.54671761, 16.50123124, 11.02734349],\n", " [ 3.66985913, 7.35171041, 5.54261877]])"]}, "execution_count": 8, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].bootstrap_err.trans_err"]}, {"cell_type": "code", "execution_count": 9, "id": "207822f9", "metadata": {}, "outputs": [{"data": {"text/plain": ["(array([0.01141016, 0.00194885, 0.00057134]),\n", " array([0.00429475, 0.00118076, 0.00095021]))"]}, "execution_count": 9, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].bootstrap_err.E_err, bdata.models[2].bootstrap_err.S_err"]}, {"cell_type": "raw", "id": "cce4f7a7", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["For E and S, the leakage/direct excitation/ :math:`\\gamma`\\ and :math:`\\beta`\\ correct values: :attr:`ModelError.Bootstrap_Error.E_err_corr ` , and :attr:`ModelError.Bootstrap_Error.S_err_corr ` "]}, {"cell_type": "code", "execution_count": 10, "id": "0a35b21a", "metadata": {}, "outputs": [{"data": {"text/plain": ["(array([0.01141016, 0.00194885, 0.00057134]),\n", " array([0.00429475, 0.00118076, 0.00095021]))"]}, "execution_count": 10, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].bootstrap_err.E_err_corr, bdata.models[2].bootstrap_err.S_err_corr"]}, {"cell_type": "raw", "id": "fa67ea1b", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["You can also see the values of the individual subset models.\n", "This is done through the attribute :attr:`ModelError.Bootstrap_Error.models ` .\n", "\n", "\n", "This attribute is again another special class, :class:`ModelError.ModelSet ` used for organizing models that vary only in their specific parameter values, but share the same number of states, and data and divisor scheme.\n", "It lets you access the transition rate, E and S values as attributes, and for E and S, access these with the :attr:`ModelError.ModelSet.trans ` , :attr:`ModelError.ModelSet.E ` , :attr:`ModelError.ModelSet.S ` attributes respectively. For E and S, the leakage/direct excitation/\\ :math:`\\gamma`\\ and :math:`\\beta`\\ correct values are accessible with the :attr:`ModelError.ModelSet.E_corr ` and :attr:`ModelError.ModelSet.S_corr ` attributes.\n", "\n", "The organization of these arrays is ``[state, subset]`` for E and S, and ``[from_state, to_state, subset]`` for transition rates. "]}, {"cell_type": "code", "execution_count": 11, "id": "944779c1", "metadata": {}, "outputs": [{"data": {"text/plain": ["array([[[1.99993025e+07, 5.47412626e+02, 1.50107894e+02],\n", " [2.56911399e+02, 1.99996697e+07, 7.33960161e+01],\n", " [2.71844461e+01, 9.25543456e+01, 1.99998803e+07]],\n", "\n", " [[1.99994454e+07, 5.01289522e+02, 5.33486484e+01],\n", " [2.02533298e+02, 1.99996915e+07, 1.06001520e+02],\n", " [7.16681031e+00, 1.38306883e+02, 1.99998545e+07]],\n", "\n", " [[1.99994312e+07, 5.68807992e+02, 2.21385758e-05],\n", " [2.15768390e+02, 1.99996411e+07, 1.43089271e+02],\n", " [4.79837339e+00, 1.25242149e+02, 1.99998700e+07]],\n", "\n", " [[1.99995343e+07, 4.36098017e+02, 2.95648463e+01],\n", " [1.55146963e+02, 1.99997508e+07, 9.40438819e+01],\n", " [7.29023492e+00, 1.10009269e+02, 1.99998827e+07]],\n", "\n", " [[1.99993033e+07, 6.12696880e+02, 8.40199625e+01],\n", " [2.11897860e+02, 1.99997083e+07, 7.98318819e+01],\n", " [8.19442843e+00, 1.01374572e+02, 1.99998904e+07]]])"]}, "execution_count": 11, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].bootstrap_err.models.trans"]}, {"cell_type": "code", "execution_count": 12, "id": "7adef573", "metadata": {"scrolled": true}, "outputs": [{"data": {"text/plain": ["(array([[0.63874832, 0.16143698, 0.06563951],\n", " [0.66284056, 0.15296923, 0.06698833],\n", " [0.67054299, 0.15811327, 0.06875795],\n", " [0.62598318, 0.16620047, 0.06867949],\n", " [0.69909413, 0.15815619, 0.06616314]]),\n", " array([[0.43073864, 0.55001657, 0.97045621],\n", " [0.43598782, 0.55806882, 0.97225223],\n", " [0.43725447, 0.55429823, 0.9687245 ],\n", " [0.43685133, 0.55273276, 0.96839454],\n", " [0.4119258 , 0.55269972, 0.97400269]]))"]}, "execution_count": 12, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].bootstrap_err.models.E, bdata.models[2].bootstrap_err.models.S"]}, {"cell_type": "code", "execution_count": 13, "id": "9ffb39d2", "metadata": {}, "outputs": [{"data": {"text/plain": ["(array([[0.63874832, 0.16143698, 0.06563951],\n", " [0.66284056, 0.15296923, 0.06698833],\n", " [0.67054299, 0.15811327, 0.06875795],\n", " [0.62598318, 0.16620047, 0.06867949],\n", " [0.69909413, 0.15815619, 0.06616314]]),\n", " array([[0.43073864, 0.55001657, 0.97045621],\n", " [0.43598782, 0.55806882, 0.97225223],\n", " [0.43725447, 0.55429823, 0.9687245 ],\n", " [0.43685133, 0.55273276, 0.96839454],\n", " [0.4119258 , 0.55269972, 0.97400269]]))"]}, "execution_count": 13, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].bootstrap_err.models.E_corr, bdata.models[2].bootstrap_err.models.S_corr"]}, {"cell_type": "raw", "id": "c9d6aecf", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": [".. _loglikerr:\n", "\n", "\n", "Loglikelihood Uncertainty Evaluation\n", "************************************\n", "\n", "\n", "\n", "The bootstrap error is very simple, however, it also can take a long time, and the particular subsets used may have a significant influence on the calculated values.\n", "\n", "What is loglikelihood uncertainty?\n", "==================================\n", "\n", "\n", "\n", "The assessment of the loglikelihood uncertainties relies on finding the loglikelihood of models where one of the model parameter values has been offset from the optimal value.\n", "While a full statistical analysis would require integration across the whole parameter space, we note that the loglikelihoods generally distribute in a Gaussian-like manner, and therefore we can approximate the uncertainty by finding the point at which the loglikelihood is some amount less than the optimal model:\n", "\n", ":math:` LL(\\lambda _{\\Delta E_{n}}) = LL(\\lambda _{optimal}) - 0.5`\\ \n", "\n", "There are two points at which this is true, one where :math:`E_{n}`\\ is greater than the optimal :math:`E_{n}`\\ , and another, where :math:`E_{n}`\\ is less than the optimal :math:`E_{n}`\\ , denoted :math:`E_{n, high}`\\ and :math:`E_{n, low}`\\ respectively. \n", "\n", "The errors reported are thus:\n", "\n", ":math:`err_{LL}(E) = \\frac{E_{n,high} - E_{n,low}}{2}`\\ \n", "\n", "Note that in most cases, :math:`E_{n, high} - E_{n, optimal} \\approx E_{n, optimal} - E_{n, low}`\\ \n", "\n", "This is not true however, for transition rates, and thus, in lieu of reporting an average value to represent a +/- type of error, instead we report directly the high and low transition rates.\n", "\n", "\n", "Calculating loglikelihood uncertainty\n", "=====================================\n", "\n", "\n", "\n", "Estimation of Loglikelihood uncertainty is handled by :class:`ModelError.Loglik_Error ` objects, which are created automatically when a :class:`H2MM_result ` object is created, and stored in the :attr:`H2MM_result.loglik_err ` attribute.\n", "Upon its creation, no values are actually calculated, only the skeleton exists.\n", "\n", "All parameter times (E/S/transition rates) follow the same basic rules, so we will start by demonstrating uncertainty estimation for E.\n", "\n", "To estimate the uncertainty for E, we use the :meth:`ModelError.Loglik_Error.get_E_err() ` method."]}, {"cell_type": "code", "execution_count": 14, "id": "45b753ce", "metadata": {}, "outputs": [{"data": {"text/plain": ["0.005624999999999991"]}, "execution_count": 14, "metadata": {}, "output_type": "execute_result"}], "source": ["E_err = bdata.models[2].loglik_err.get_E_err(0)\n", "E_err"]}, {"cell_type": "raw", "id": "5d4acd59", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["The equivalent for the stoichiometry is :meth:`ModelError.Loglik_Error.get_S_err() ` :"]}, {"cell_type": "code", "execution_count": 15, "id": "b498cd4b", "metadata": {"scrolled": false}, "outputs": [{"data": {"text/plain": ["0.00077636718749996"]}, "execution_count": 15, "metadata": {}, "output_type": "execute_result"}], "source": ["S_err = bdata.models[2].loglik_err.get_S_err(2)\n", "S_err"]}, {"cell_type": "raw", "id": "f313e780", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["These value indicate the point at which models where a given parameter value is varied from the optimal, have a loglikelihood 0.5 less than the optimal model, ie: \n", "\n", ":math:` LL(\\lambda _{\\Delta E_{n}}) = LL(\\lambda _{optimal}) - 0.5`\\ \n", "\n", "There are two points at which this is true, one where :math:`E_{n}`\\ is greater than the optimal :math:`E_{n}`\\ , and another, where :math:`E_{n}`\\ is less than the optimal :math:`E_{n}`\\ , denoted :math:`E_{n, high}`\\ and :math:`E_{n, low}`\\ respectively. \n", "\n", "The errors reported are thus:\n", "\n", ":math:`err_{LL}(E) = \\frac{E_{n,high} - E_{n,low}}{2}`\\ \n", "\n", "Note that in most cases, :math:`E_{n, high} - E_{n, optimal} \\approx E_{n, optimal} - E_{n, low}`\\ \n", "\n", "This is not true however, for transition rates, and thus, in lieu of reporting an average value to represent a +/- type of error, instead we report directly the high and low transition rates."]}, {"cell_type": "code", "execution_count": 16, "id": "5f310745", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(data=[501.8588059030635, 562.3571669228339],\n", " mask=[False, False],\n", " fill_value=inf)"]}, "execution_count": 16, "metadata": {}, "output_type": "execute_result"}], "source": ["trans_err = bdata.models[2].loglik_err.get_trans_err(0,1)\n", "trans_err"]}, {"cell_type": "raw", "id": "d725520b", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["The :attr:`ModelError.Loglik_Error.get_E/S_err() ` methods also allow passing the keyword parameter ``simple`` as ``simple=False`` to return the low/high values like :meth:`ModelError.Loglik_Error.get_trans_err() ` "]}, {"cell_type": "code", "execution_count": 17, "id": "7b9ba05d", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(data=[0.6546853599116289, 0.6659353599116289],\n", " mask=[False, False],\n", " fill_value=nan)"]}, "execution_count": 17, "metadata": {}, "output_type": "execute_result"}], "source": ["E_err = bdata.models[2].loglik_err.get_E_err(0, simple=False)\n", "E_err"]}, {"cell_type": "code", "execution_count": 18, "id": "d3ea4678", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(data=[0.4273356364488786, 0.43415204269887864],\n", " mask=[False, False],\n", " fill_value=nan)"]}, "execution_count": 18, "metadata": {}, "output_type": "execute_result"}], "source": ["S_err = bdata.models[2].loglik_err.get_S_err(0, simple=False)\n", "S_err"]}, {"cell_type": "raw", "id": "ee30b991", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Accessing Values Previously Calculated\n", "======================================\n", "\n", "\n", "\n", "Calculations of the uncertainty values are stored in masked arrays, so that only calculated values are available.\n", "These can be accessed through the attributes :attr:`ModelError.Loglik_Error.E ` , :attr:`ModelError.Loglik_Error.S ` , :attr:`ModelError.Loglik_Error.trans ` ."]}, {"cell_type": "code", "execution_count": 19, "id": "9a5f2313", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(\n", " data=[[[--, --],\n", " [501.8588059030635, 562.3571669228339],\n", " [--, --]],\n", "\n", " [[--, --],\n", " [--, --],\n", " [--, --]],\n", "\n", " [[--, --],\n", " [--, --],\n", " [--, --]]],\n", " mask=[[[ True, True],\n", " [False, False],\n", " [ True, True]],\n", "\n", " [[ True, True],\n", " [ True, True],\n", " [ True, True]],\n", "\n", " [[ True, True],\n", " [ True, True],\n", " [ True, True]]],\n", " fill_value=inf)"]}, "execution_count": 19, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.trans"]}, {"cell_type": "code", "execution_count": 20, "id": "0a674c6d", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(data=[0.005624999999999991, --, --],\n", " mask=[False, True, True],\n", " fill_value=nan)"]}, "execution_count": 20, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.E"]}, {"cell_type": "code", "execution_count": 21, "id": "d50bc959", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(data=[0.003408203125000009, --, 0.00077636718749996],\n", " mask=[False, True, False],\n", " fill_value=nan)"]}, "execution_count": 21, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.S"]}, {"cell_type": "raw", "id": "4ba20909", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["The :attr:`ModelError.Loglik_Error.E ` and :attr:`ModelError.Loglik_Error.S ` are a little more processed than the :attr:`ModelError.Loglik_Error.trans ` , as these do not show the low/high directly for a given state, but rather show half the difference between them.\n", "\n", "\n", "If you want to see the actual low and high values, these can be accessed with the :attr:`ModelError.Loglik_Error.E_lh ` and :attr:`ModelError.Loglik_Error.S_lh ` attributes (this is basically the same as passing the ``simple=False`` keyword argument to :meth:`ModelError.Loglik_Error.get_E_err() ` / :meth:`ModelError.Loglik_Error.get_S_err() ` ):"]}, {"cell_type": "code", "execution_count": 22, "id": "d50f1b6d", "metadata": {"scrolled": true}, "outputs": [{"data": {"text/plain": ["masked_array(\n", " data=[[0.6546853599116289, 0.6659353599116289],\n", " [--, --],\n", " [--, --]],\n", " mask=[[False, False],\n", " [ True, True],\n", " [ True, True]],\n", " fill_value=nan)"]}, "execution_count": 22, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.E_lh"]}, {"cell_type": "code", "execution_count": 23, "id": "85533278", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(\n", " data=[[0.4273356364488786, 0.43415204269887864],\n", " [--, --],\n", " [0.9700226579124809, 0.9715753922874808]],\n", " mask=[[False, False],\n", " [ True, True],\n", " [False, False]],\n", " fill_value=nan)"]}, "execution_count": 23, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.S_lh"]}, {"cell_type": "raw", "id": "a93613e3", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Concluding this list of access attributes, the loglikelihood values of these models are also stored in the :attr:`ModelError.Loglik_Error.E_ll ` , :attr:`ModelError.Loglik_Error.S_ll ` , and :attr:`ModelError.Loglik_Error.trans_ll ` attributes:"]}, {"cell_type": "code", "execution_count": 24, "id": "5ccdb91f", "metadata": {"scrolled": true}, "outputs": [{"data": {"text/plain": ["masked_array(\n", " data=[[-133208.37002081441, -133208.37125588267],\n", " [--, --],\n", " [--, --]],\n", " mask=[[False, False],\n", " [ True, True],\n", " [ True, True]],\n", " fill_value=-inf)"]}, "execution_count": 24, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.E_ll"]}, {"cell_type": "code", "execution_count": 25, "id": "00df5594", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(\n", " data=[[-133208.36515007372, -133208.36806724593],\n", " [--, --],\n", " [-133208.36725677058, -133208.36851899815]],\n", " mask=[[False, False],\n", " [ True, True],\n", " [False, False]],\n", " fill_value=-inf)"]}, "execution_count": 25, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.S_ll"]}, {"cell_type": "raw", "id": "1404bab1", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Adjusting thresholds\n", "====================\n", "\n", "\n", "\n", "From the earlier equation, the estimated error is defined by having a loglikelihood :math:`0.5`\\ less than the optimal, however, if you wish to change this threshold, to say :math:`1.0`\\ , this can be done (before running any ``get_`` method) by setting the :attr:`ModelError.Loglik_Error.thresh ` attribute:"]}, {"cell_type": "code", "execution_count": 26, "id": "23d0f868", "metadata": {}, "outputs": [{"data": {"text/plain": ["0.00277343749999992"]}, "execution_count": 26, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.thresh = 1.0\n", "S_err = bdata.models[2].loglik_err.get_S_err(1)\n", "S_err"]}, {"cell_type": "raw", "id": "82b8c950", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Another factor that can be adjusted is how precisely the search algorithm needs to find the loglikelihood, this factor is stored in the :attr:`ModelError.Loglik_Error.flex ` attribute. The default is :math:`0.005`\\ "]}, {"cell_type": "code", "execution_count": 27, "id": "0b2927d8", "metadata": {}, "outputs": [{"data": {"text/plain": ["0.002695312500000005"]}, "execution_count": 27, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.flex = 5e-2\n", "E_err = bdata.models[2].loglik_err.get_E_err(1)\n", "E_err"]}, {"cell_type": "raw", "id": "5d0d4179", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["These set univeral threshold/flex values, is generally preferred, these values can be altered for each calculation.\n", "This works by passing ``thresh`` and ``flex`` keyword arguments to the :meth:`ModelError.Loglik_Error.get_E_err() ` / :meth:`ModelError.Loglik_Error.get_S_err() ` / :meth:`ModelError.Loglik_Error.get_trans_err() ` .\n", "\n", ".. warning::\n", "\n", " Passing ``thresh`` and ``flex`` keyword arguments to the :meth:`ModelError.Loglik_Error.get_E_err() ` / :meth:`ModelError.Loglik_Error.get_S_err() ` / :meth:`ModelError.Loglik_Error.get_trans_err() ` will only affect the current calculation.\n", " Therefore all other calculations will have different ``thresh`` and ``flex`` values, and therefore will not be comparable to one another.\n", " This is why it is discouraged to use this method.\n", "\n", ""]}, {"cell_type": "code", "execution_count": 28, "id": "ad55f2f2", "metadata": {}, "outputs": [{"data": {"text/plain": ["0.0010156249999999922"]}, "execution_count": 28, "metadata": {}, "output_type": "execute_result"}], "source": ["E_err = bdata.models[3].loglik_err.get_E_err(2, thresh=0.1, flex=1e-2)\n", "E_err"]}, {"cell_type": "raw", "id": "bc5dc49b", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Clearing Values\n", "===============\n", "\n", "\n", "\n", "Since values already stored are not recalculated, if a new threshold is set, previous values can be cleared using the :meth:`ModelError.Loglik_Error.clear_E ` :meth:`ModelError.Loglik_Error.clear_S() ` , and :meth:`ModelError.Loglik_Error.clear_trans() ` , and :meth:`ModelError.Loglik_Error.clear_all() ` methods.\n", "These reset their respective arrays:"]}, {"cell_type": "code", "execution_count": 29, "id": "a16837a5", "metadata": {}, "outputs": [], "source": ["bdata.models[2].loglik_err.clear_E()\n", "bdata.models[2].loglik_err.clear_S()\n", "bdata.models[2].loglik_err.clear_trans()"]}, {"cell_type": "raw", "id": "71836f1b", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["As the name suggests, :meth:`ModelError.Loglik_Error.clear_all() ` clears all the values.\n", "So the three lines above, together do what is done bellow in a single line:"]}, {"cell_type": "code", "execution_count": 30, "id": "24329d56", "metadata": {}, "outputs": [], "source": ["bdata.models[2].loglik_err.clear_all()"]}, {"cell_type": "raw", "id": "608dce62", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Calculating All Uncertainty Values\n", "==================================\n", "\n", "\n", "\n", "When a given :class:`H2MM_result ` has only a few states, it will not take long to characterize the uncertainty of all model parameters, but for 5+ states, this becomes time consuming.\n", "Hence the choice to break with burstH2MM's normal strategy of calculating on demand and storing the result, and instead using a :meth:`ModelError.Loglik_Error.get_E_err ` / :meth:`ModelError.Loglik_Error.get_S_err ` / :meth:`ModelError.Loglik_Error.get_trans_err ` strategy, and masked arrays, that unmask values that have been calculated.\n", "\n", "However, if you so desire, the :meth:`ModelError.Loglik_Error.all_eval() ` method provides a shortcut, and evaluates all parameters for you.\n", "This is great for 2 and 3 state models, workable for 4 state models, and less advisable for 5+ state models."]}, {"cell_type": "code", "execution_count": 31, "id": "9b1b647a", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(\n", " data=[[[--, --],\n", " [501.8588059030635, 562.3571669228339],\n", " [40.92595696232948, 67.31424834784906]],\n", "\n", " [[193.29419084215726, 217.69955433877388],\n", " [--, --],\n", " [95.36877509434665, 111.5681913319069]],\n", "\n", " [[3.735882210708648, 12.60135444968205],\n", " [108.29230392786711, 124.56154227875899],\n", " [--, --]]],\n", " mask=[[[ True, True],\n", " [False, False],\n", " [False, False]],\n", "\n", " [[False, False],\n", " [ True, True],\n", " [False, False]],\n", "\n", " [[False, False],\n", " [False, False],\n", " [ True, True]]],\n", " fill_value=inf)"]}, "execution_count": 31, "metadata": {}, "output_type": "execute_result"}], "source": ["# reset thresh and flex to default before re-calculating everything\n", "bdata.models[2].loglik_err.thresh, bdata.models[2].loglik_err.flex = 0.5, 5e-3\n", "\n", "# evaluate all parameters\n", "bdata.models[2].loglik_err.all_eval()\n", "bdata.models[2].loglik_err.trans"]}, {"cell_type": "code", "execution_count": 32, "id": "7e4ccd88", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(data=[0.005624999999999991, 0.0019238281250000128,\n", " 0.0010351562500000022],\n", " mask=[False, False, False],\n", " fill_value=nan)"]}, "execution_count": 32, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.E"]}, {"cell_type": "code", "execution_count": 33, "id": "dae5aa06", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(data=[0.003408203125000009, 0.001953125, 0.00077636718749996],\n", " mask=[False, False, False],\n", " fill_value=-inf)"]}, "execution_count": 33, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.S"]}, {"cell_type": "raw", "id": "99d7195c", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Examining Loglikelihood Variance Along a Given Parameter\n", "========================================================\n", "\n", "\n", "\n", "So far uncertainty estimation has been built finding how far a given parameter must be varied to alter the loglikelihood by a certain value.\n", "This is done by an iterative process, and each iteration is saved.\n", "The parameter values and loglikelihoods are stored in attributes of :class:`ModelError.Loglik_Error ` .\n", "\n", "+------------------+---------------------------------------------------------------------------------------------+-----------------------------------------------------------------------------------------+\n", "| Parameter type | Parameter Value | Loglikelihood |\n", "+==================+=============================================================================================+=========================================================================================+\n", "| E | \\ :attr:`ModelError.Loglik_Error.E_rng ` | \\ :attr:`ModelError.Loglik_Error.E_ll_rng ` |\n", "+------------------+---------------------------------------------------------------------------------------------+-----------------------------------------------------------------------------------------+\n", "| S | \\ :attr:`ModelError.Loglik_Error.S_rng ` | \\ :attr:`ModelError.Loglik_Error.S_ll_rng ` |\n", "+------------------+---------------------------------------------------------------------------------------------+-----------------------------------------------------------------------------------------+\n", "| Trans | \\ :attr:`ModelError.Loglik_Error.t_rate_rng ` | \\ :attr:`ModelError.Loglik_Error.t_ll_rng ` |\n", "+------------------+---------------------------------------------------------------------------------------------+-----------------------------------------------------------------------------------------+\n", "\n", "\n", "\n", "These are 2D (for trans) and 1D (for E/S) numpy object arrays, whose elements are 1D numpy arrays of all values/loglikelihoods evaluates thus far.\n", "\n", "These arrays however will often have values clustered around where the search value is, whereas for more in-depth analysis, it would be more convenient to have an evenly spaced set of values.\n", "Such arrays can be generated by the :meth:`ModelError.Loglik_Error.E_space() ` , :meth:`ModelError.Loglik_Error.S_space() ` , and :meth:`ModelError.Loglik_Error.trans_space() ` methods.\n", "\n", "The only required argument to these functions is the state or transition for which to calculate the array.\n", "The ``rng`` and ``steps`` keyword arguments let you specify the range over which to space the parameter values, and the number of values to evaluate (think of this as similar to the numpy linspace and logspace functions).\n", "\n", "``steps`` is always a positive integer.\n", "\n", "``rng`` however has several options:\n", "\n", "+------------------+-----------------------------------+------------------+------------------+\n", "| Option | Behavior | When callable | \\ ``steps`` |\n", "| | | | ignored |\n", "+==================+===================================+==================+==================+\n", "| int/float | multiply value by difference | Must have | No |\n", "| | between low/high value and | evaluated | |\n", "| | optimal value to offset for | uncertainty | |\n", "| | low/high values of range | before calling | |\n", "+------------------+-----------------------------------+------------------+------------------+\n", "| 2 element | low/high values for the range | Call anytime | No |\n", "| array-like | | | |\n", "| (tuple, list, | | | |\n", "| numpy array) | | | |\n", "+------------------+-----------------------------------+------------------+------------------+\n", "| Many element | The individual values of the | Call anytime | Yes |\n", "| array-like | specified parameter to evaluate | | |\n", "| | the matrix, steps ignored | | |\n", "+------------------+-----------------------------------+------------------+------------------+\n", "\n", "\n", ".. note::\n", "\n", " If ``rng`` is not specified, it behaves like ``rng=2`` and therefore prior to evaluating, you must have already performed :meth:`ModelError.Loglik_Error.get_E_err() ` / :meth:`ModelError.Loglik_Error.get_S_err() ` / :meth:`ModelError.Loglik_Error.get_trans_err() ` for the given state or transition.\n", "\n", ""]}, {"cell_type": "code", "execution_count": 34, "id": "ec7d30e2", "metadata": {}, "outputs": [{"data": {"text/plain": ["(0.6490603599116289, 0.6715603599116289, 20)"]}, "execution_count": 34, "metadata": {}, "output_type": "execute_result"}], "source": ["Erng, Ell = bdata.models[2].loglik_err.E_space(0)\n", "Erng[0], Erng[-1], Erng.size"]}, {"cell_type": "code", "execution_count": 35, "id": "8326eb41", "metadata": {}, "outputs": [{"data": {"text/plain": ["(0.5476305043691996, 0.5593492543691996, 10)"]}, "execution_count": 35, "metadata": {}, "output_type": "execute_result"}], "source": ["Srng, Sll = bdata.models[2].loglik_err.S_space(1, rng=3, steps=10)\n", "Srng[0], Srng[-1], Srng.size"]}, {"cell_type": "raw", "id": "db5296ee", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": [".. note::\n", "\n", " For :meth:`ModelError.Loglik_Error.trans_space() ` the transition must be specified as a 2-tuple of (from_state, to_state).\n", "\n", ""]}, {"cell_type": "code", "execution_count": 36, "id": "9050f73c", "metadata": {}, "outputs": [{"data": {"text/plain": ["(99.99999999999999, 400.0, (20,))"]}, "execution_count": 36, "metadata": {}, "output_type": "execute_result"}], "source": ["trng, tll = bdata.models[2].loglik_err.trans_space((0,1), rng=(100, 400))\n", "trng[0], trng[-1], trng.shape"]}, {"cell_type": "raw", "id": "579d2b0b", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["The return values of these :meth:`ModelError.Loglik_Error.E_space() ` / :meth:`ModelError.Loglik_Error.S_space() ` / :meth:`ModelError.Loglik_Error.trans_space() ` methods are automatically added to the :attr:`ModelError.Loglik_Error.E_rate_rng ` / :attr:`ModelError.Loglik_Error.S_rate_rng ` / :attr:`ModelError.Loglik_Error.t_rate_rng ` and :attr:`ModelError.Loglik_Error.E_ll_rng ` / :attr:`ModelError.Loglik_Error.S_ll_rng ` / :attr:`ModelError.Loglik_Error.t_ll_rng ` attributes, and so we can access all previously calculated values.\n", "Note that these are stored in arrays of arrays, so you must specify the state/(from_state, to_state) that you want to access."]}, {"cell_type": "code", "execution_count": 37, "id": "19680c3c", "metadata": {}, "outputs": [{"data": {"text/plain": ["array([-133421.33099232, -133408.78806882, -133396.23640999,\n", " -133383.70003837, -133371.20575538, -133358.78339051,\n", " -133346.46606703, -133334.29048486, -133322.29722086,\n", " -133310.53104671, -133299.04126446, -133287.88205962,\n", " -133277.11287135, -133266.79877898, -133257.01090395,\n", " -133247.82682552, -133239.33100829, -133231.61523905,\n", " -133224.77906936, -133218.93026009, -133209.34821107,\n", " -133208.70964879, -133208.45556987, -133208.39896862,\n", " -133208.37171389, -133208.34515839, -133208.24595844,\n", " -133207.99849657, -133208.16408984, -133208.27247109,\n", " -133208.33312831, -133208.36508243, -133208.3981236 ])"]}, "execution_count": 37, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.t_ll_rng[0,1]"]}, {"cell_type": "raw", "id": "e49bde2e", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Plotting Parameters\n", "===================\n", "\n", "\n", "\n", "It is also possible to plot how the loglikelihood varies along a given parameter.\n", "\n", "This uses the :func:`ll_E_scatter() ` , :func:`ll_S_scatter() ` and :func:`ll_trans_scatter() ` functions.\n", "These functions take as required arguments a :class:`H2MM_result ` or :class:`ModelError.Loglik_Error ` object, and which state to plot. For :func:`ll_E_scatter() ` and :func:`ll_S_scatter ` , this is a single integer, indicating the index of the state. For :func:`ll_trans_scatter() ` this is two arguments: ``from_state`` and ``to_state`` ."]}, {"cell_type": "code", "execution_count": 38, "id": "9217efa5", "metadata": {}, "outputs": [{"data": {"text/plain": [""]}, "execution_count": 38, "metadata": {}, "output_type": "execute_result"}, {"data": {"image/png": "iVBORw0KGgoAAAANSUhEUgAAAZAAAAEVCAYAAADOwrOnAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAf1ElEQVR4nO3df5BdZZ3n8feH/NAeJHQCBNJNmKAVekSyJNJSKDuC/ArELRJgsGAtSS1qHBQV1kltslCMLOUACY6jta4sAzixdPxJCAHRJkQZlilAOiQkQWgSWBzSnYWoNOjYSojf/eM8F26u9/aP0337nu7+vKpOnec+z3PO/d6G3O85z3PuOYoIzMzMhuqARgdgZmZjkxOImZnl4gRiZma5OIGYmVkuTiBmZpaLE4iZmeXiBFJG0oWSnpT0R0ntNfq8VdLPJD2R+l5b1nadpK2Stki6T1JLqj9T0iZJ29L6tLJtTkj1OyV9RZJS/VskfTfVPyppTtk2SyXtSMvSsvqjU98dadupqV5p3ztTfO8u2+ZsSV2pbUVZ/QxJG9K+NkiaPoi/37702bdIWj/IP7uZjVUR4SUtwDuBNuABoL1GHwFvS+UpwKPASen1tLJ+nwFuTuUFQEsqHwd0l/X7GfDetN8fAeek+k+WbX8R8N1UngE8l9bTU3l6avsecFEq3wxclsqL0r4FnAQ8muonAc8CbwemAk8Ax6a2VcCKVF4B3DiIv99vG/3f0IsXL6O3+AykTEQ8FRFdA/SJiPhtejklLZHaXi3remBZ/eaI6En1TwJvTWcYs8iSzsMREcA3gCWp32JgTSr/ADg9nZ0sBDZExK8j4mVgA3B2ajst9SVtW76vb6TYHwGa03ufCOyMiOci4jXgO6lv5fu/sS9JkyStlvRYOpv5RH9/LzMbv5xAckhfoluAl8i+zB8ta/uCpBeADwPXVNn8AmBzRPwBaAV2lbXtSnWk9QsAEfE68ApwSHl9xTaHAL2pb819VbTVqgc4PCJ2p/ffDcxM9R8FXomI9wDvAT4u6ejU9lZJnZIekbSkymc3s3FkwiUQSfdL2l5lWTzw1pmI2BcR84EjgRMlHVfWdlVEzAa+BVxe8d7vAm4ESkftqrb7AdqGWp9nX/05C7gkJdBHyRLX3NR2VES0A/8Z+AdJ7xhgX2Y2hk1udACjLSLOGMF99Up6ADgb2F7R/M/AD4G/BZB0JHAncElEPJv67CJLQiVHAj1lbbOBXZImAwcDv071p1Zs8wDwS7KhqcnpLKTavirfZ2qNeoAXJc2KiN1puOulVC/g0xHRUeXv0ZPWz6W/ywKyORYzG4cm3BnIcEk6TFJzKjcBZwBPp9dzy7qeW1bfTJZMVkbEv5Y6pKGh30g6Kc1hXALclZrXA6UrrP4K+EmaJ+kAzpI0PV0ZdRbQkdp+mvqSti3f1yXpaqyTyIagdgOPAXPT1VtTySbr15dts7TKvjqAyyRNSZ/tGEkHpnjekuoOBU4Gfj7Yv6uZjUGNnsUv0gKcR3a0/gfgRbIvZoAW4N5U/g/AZmAr2VnHNWXb35HqtgJ3A62p/mrg34EtZcvM1NaetnkW+J+AUv1bge8DO8mu1Hp72ftcmup3Av+lrP7tqe/OtO1bUr2Ar6b32EbZFWZkV2g9k9quKqs/BNgI7EjrGan+AODv0n62kyWtg4H3pbon0vqjjf7v6cWLl/oupS8rMzOzIfEQlpmZ5eIEYmZmuUyoq7AOPfTQmDNnTqPDMDMbUzZt2vTLiDissn5CJZA5c+bQ2dnZ6DDMzMYUSb+oVu8hLDMzy8UJxMzMcnECMTOzXJxAzMwslwk1iW422tZt7mZ1Rxc9vX20NDexfGEbSxa0Dryh2RjgBGJWJ+s2d7Ny7Tb69u4DoLu3j5VrtwEMOok4AVmReQjLrE5Wd3S9kTxK+vbuY3VHv88se0MpAXX39hG8mYDWbe6uQ7RmQ+cEYlYnPb19Q6qvNNwEZFZvTiBmddLS3DSk+krDTUBm9eY5ELN+DGcOYvnCtv3mQACapkxi+cK2QW3f0txEd5VkMdgEBJ5DsfryGYhZDcOdg1iyoJXrz59Ha3MTAlqbm7j+/HlDSkBNUybtVzeUBOQ5FKs3n4GY1dDfHMRgk8CSBa25j/hL2+U9gxiJ+M364wRiVkMR5iCGk4CKEL+Nbx7CMqthuJPgjTbW47ficwIxq2G4cxCNNtbjt+LzEJZZDcOdg2i0sR6/FZ8iotExjJr29vbwA6XMzIZG0qaIaK+sb8gQlqT5kh6RtEVSp6QTa/Q7W1KXpJ2SVpTVz5C0QdKOtJ4+etHbWLJuczcn3/ATjl7xQ06+4Se+hNVsBDVqDmQVcG1EzAeuSa/3I2kS8FXgHOBY4GJJx6bmFcDGiJgLbEyvzfbj30EMnxOw9adRCSSAaal8MNBTpc+JwM6IeC4iXgO+AyxObYuBNam8BlhSv1BtrPK9pIbHCdgG0qgEcgWwWtILwE3Ayip9WoEXyl7vSnUAh0fEboC0nlnrjSQtS8NknXv27BmJ2G2M8O8ghscJ2AZStwQi6X5J26ssi4HLgCsjYjZwJXBbtV1UqRvyjH9E3BIR7RHRfthhhw11cxvD/DuI4XECtoHULYFExBkRcVyV5S5gKbA2df0+2XBVpV3A7LLXR/LmUNeLkmYBpPVL9fkUNpb5dxDD4wRsA2nUEFYPcEoqnwbsqNLnMWCupKMlTQUuAtantvVkSYi0vquOsdoYNdybGU50TsA2kEb9kPDjwJclTQZ+DywDkNQC3BoRiyLidUmXAx3AJOD2iHgybX8D8D1JHwX+Dbhw1D+BjQnDuZfUROcfItpA/ENCMzPrV6F+SGhmZmOfE4iZmeXiBGJmZrk4gZiZWS6+nbsV2rrN3b4KyKygnECssEr3YirdTqN0LybAScSsAJxArLD6uxeTE8jY4DPI8c0JxArL92Ia23wGOf55Et0Ky/diGtt8N9/xzwnECsv3YhrbfAY5/jmBWGH5Zohjm88gxz/PgVih+WaIY9fyhW37zYGAzyDHGycQM6sL3813/HMCMbO68Rnk+OY5EDMzy6UhCUTSfEmPSNoiqVPSnzzSVtJsST+V9JSkJyV9tqzt85K60/ZbJC0a3U9gZmaNGsJaBVwbET9KX/6rgFMr+rwOfC4iHpd0ELBJ0oaI+Hlq/1JE3DR6IZuZWblGDWEFMC2VDyZ7Rvr+HSJ2R8Tjqfwb4CnAg6lmZgXRqDOQK4AOSTeRJbH39ddZ0hxgAfBoWfXlki4BOsnOVF6uT6hmZlZN3c5AJN0vaXuVZTFwGXBlRMwGrgRu62c/bwPuAK6IiFdT9deAdwDzgd3AF/vZflmaZ+ncs2fPyHw4MzNDETH6byq9AjRHREgS8EpETKvSbwpwD9AREX9fY19zgHsi4riB3re9vT06OzuHF7yZ2QQjaVNEtFfWN2oOpAc4JZVPA3ZUdkiJ5TbgqcrkIWlW2cvzgO11itPMzGpo1BzIx4EvS5oM/B5YBiCpBbg1IhYBJwMfAbZJ2pK2++8RcS+wStJ8ssn454FPjGr0Nmh+HoTZ+NWQBBIRDwEnVKnvARaV9VGN7T9S1wBtRPh5EDZcPgApNv8S3erGz4Ow4SgdgHT39hG8eQCybnN3o0OzxAnE6sbPg7Dh8AFI8TmBWN34eRA2HD4AKT4nEKsbP1HQhsMHIMXnBGJ14ycK2nD4AKT4/DwQqys/D8Ly8gOpis8JxMwKywcgxeYhLDMzy8UJxMzMcnECMTOzXJxAzMwsFycQMzPLxQnEzMxycQIxM7NcnEDMzCwXJxAzM8ulIQlE0nxJj0jaIqlT0ok1+j0vaVupX1n9DEkbJO1I6+mjF72ZmUHjzkBWAddGxHzgmvS6lg9ExPyKB7qvADZGxFxgY3ptZmajqFEJJIBpqXww0DPE7RcDa1J5DbBkZMIyM7PBatTNFK8AOiTdRJbE3lejXwD3SQrgf0fELan+8IjYDRARuyXNrHfAZma2v7olEEn3A0dUaboKOB24MiLukPQh4DbgjCp9T46InpQgNkh6OiIeHGIcy4BlAEcdddSQPoNlz6X27bTNrBpFxOi/qfQK0BwRIUnAKxExbYBtPg/8NiJuktQFnJrOPmYBD0TEgE+ZaW9vj87OzoG6WbJuczcr127b77nUTVMm+aFQNmb4AGhkSNpUMQ8NNG4OpAc4JZVPA3ZUdpB0oKSDSmXgLGB7al4PLE3lpcBddY12glrd0bVf8gDo27uP1R1dDYrIbPBKB0DdvX0E0N3bx8q121i3ubvRoY0bjUogHwe+KOkJ4O9IQ0ySWiTdm/ocDjyU+vwM+GFE/Di13QCcKWkHcGZ6bSOsp7dvSPVmReIDoPpryCR6RDwEnFClvgdYlMrPAcfX2P5XZPMoVkctzU10V0kWLc1NDYjGbGh8AFR//iW61bR8YRtNUybtV9c0ZRLLFw443WTWcLUOdHwANHKcQKymJQtauf78ebQ2NyGgtbnJE+g2ZvgAqP4a9TsQGyOWLGh1wrAxqfT/ra/Cqh8nEDMbt3wAVF8ewjIzs1ycQMzMLBcnEDMzy8UJxMzMcnECMTOzXJxAzMwsFycQMzPLxQnEzMxycQIxM7NcnEDMzCwXJxAzM8vFCcTMzHJpSAKRNF/SI5K2SOqUdGKVPm2pvbS8KumK1PZ5Sd1lbYtG/UOYmU1wjbob7yrg2oj4UfryXwWcWt4hIrqA+QCSJgHdwJ1lXb4UETeNSrRmZvYnGjWEFcC0VD4Y6Bmg/+nAsxHxi7pGZWZmg9aoM5ArgA5JN5ElsfcN0P8i4NsVdZdLugToBD4XES9X21DSMmAZwFFHHTWcmM3MrIwioj47lu4HjqjSdBXZGcW/RMQdkj4ELIuIM2rsZyrZGcq7IuLFVHc48EuyM5nrgFkRcelAMbW3t0dnZ2euz2NmNlFJ2hQR7ZX1dTsDqZUQUjDfAD6bXn4fuLWfXZ0DPF5KHmnfb5Ql/SNwz/CiHb/Wbe72Iz3NrC4aNYTVA5wCPACcBuzop+/FVAxfSZoVEbvTy/OA7XWIccxbt7mblWu30bd3HwDdvX2sXLsNwEnEbBB8ANa/RiWQjwNfljQZ+D1pjkJSC3BrRCxKr/8MOBP4RMX2qyTNJxvCer5KuwGrO7reSB4lfXv3sbqjy/8IzAbgA7CBNSSBRMRDwAlV6nuARWWvfwccUqXfR+oa4DjR09s3pHoze5MPwAbmX6KPYy3NTUOqN7M3+QBsYE4g49jyhW00TZm0X13TlEksX9jWoIjMxg4fgA3MCWQcW7KglevPn0drcxMCWpubuP78eT79NhsEH4ANLPcciKR/iwj/Mq/glixodcIwy6H078ZXYdU2nEl0jVgUZmYF5AOw/g1nCKs+P2E3M7Mxod8zEEn/tVYT8LaRD8fMzMaKgYawDuqn7csjGYiZmY0t/SaQiLi2Vlvp4U5mZjYxDWcOpNbwlpmZTQDDSSC+CsvMbALzVVhmZpbLQFdh/YbqiUKAf89vZjaBDTSJ3t9VWGZmNoH5XlhmZpaLE4iZmeXSkAQi6XhJD0vaJuluSdNq9DtbUpeknZJWlNXPkLRB0o60nj560ZuZGTTuDORWYEVEzAPuBJZXdpA0CfgqcA5wLHCxpGNT8wpgY0TMBTam12ZmNooalUDagAdTeQNwQZU+JwI7I+K5iHgN+A6wOLUtBtak8hpgSf1CNTOzahqVQLYD56byhcDsKn1agRfKXu9KdQCHR8RugLSeWeuNJC2T1Cmpc8+ePcMO3MzMMnVLIJLul7S9yrIYuBT4lKRNZDdsfK3aLqrUDfnHixFxS0S0R0T7YYcdNtTNzcyshuE8UKpfEXHGAF3OApB0DPDBKu272P/M5EigJ5VflDQrInZLmgW8NNx4zcxsaBp1FdbMtD4AuBq4uUq3x4C5ko6WNBW4CFif2tYDS1N5KXBXfSM2M7NKjZoDuVjSM8DTZGcVXweQ1CLpXoCIeB24HOgAngK+FxFPpu1vAM6UtAM4M702M7NRpIiJc0/E9vb26OzsbHQYZmZjiqRNEdFeWe9fopuZWS5OIGZmlkvdrsKykbFuczerO7ro6e2jpbmJ5QvbWLKgdeANzczqzAmkwNZt7mbl2m307d0HQHdvHyvXbgNwEjEbA8b7AaCHsApsdUfXG8mjpG/vPlZ3dDUoIjMbrNIBYHdvH8GbB4DrNnc3OrQR4wRSYD29fUOqN7PimAgHgE4gBdbSXP2pwbXqzaw4JsIBoBNIgS1f2EbTlEn71TVNmcTyhW0NisjMBmsiHAA6gRTYkgWtXH/+PFqbmxDQ2tzE9efPG1eTcGbj1UQ4APRVWAW3ZEGrE4bZGFT6dzuer8JyAjEzq5PxfgDoISwzM8vFCcTMzHJxAjEzs1ycQMzMLJdGPZHweEkPS9om6W5J06r0mS3pp5KekvSkpM+WtX1eUrekLWlZNLqfwMzMGnUGciuwIiLmAXcCy6v0eR34XES8EzgJ+JSkY8vavxQR89Nyb/1DNjOzco1KIG3Ag6m8AbigskNE7I6Ix1P5N2SPtR2/18OZmY0xjUog24FzU/lCYHZ/nSXNARYAj5ZVXy5pq6TbJU2vS5RmZlZT3RKIpPslba+yLAYuJRuS2gQcBLzWz37eBtwBXBERr6bqrwHvAOYDu4Ev9rP9Mkmdkjr37NkzMh/OzMxQRDQ2AOkY4JsRcWKVtinAPUBHRPx9je3nAPdExHEDvVd7e3t0dnYOM2Izs4lF0qaIaK+sb9RVWDPT+gDgauDmKn0E3AY8VZk8JM0qe3ke2ZCYmZmNokbNgVws6RngaaAH+DqApBZJpSuqTgY+ApxW5XLdVekS4K3AB4ArRzl+M7MJr+FDWKPJQ1hmZkNXqCEsMzMb+5xAzMwsFycQMzPLxQnEzMxycQIxM7NcnEDMzCwXJxAzM8vFCcTMzHJxAjEzs1ycQMzMLBcnEDMzy8UJxMzMcnECMTOzXCY3OoDxbt3mblZ3dNHT20dLcxPLF7axZIEf7W5mY58TSB2t29zNyrXb6Nu7D4Du3j5Wrt0G4CRiZoNS5INQD2HV0eqOrjeSR0nf3n2s7uhqUERmNpaUDkK7e/sI3jwIXbe5u9GhAY17pO3xkh5OTxW8W9K0Gv2eT322SOosq58haYOkHWk9ffSiH7ye3r4h1ZuZlSv6QWijzkBuBVZExDzgTmB5P30/EBHzK56GtQLYGBFzgY3pdeG0NDcNqd7MrFzRD0IblUDagAdTeQNwwRC3XwysSeU1wJKRCWtkLV/YRtOUSfvVNU2ZxPKFbQ2KyMzGkqIfhDYqgWwHzk3lC4HZNfoFcJ+kTZKWldUfHhG7AdJ6Zq03krRMUqekzj179oxA6IO3ZEEr158/j9bmJgS0Njdx/fnzCjMBZmbFVvSDUEVEfXYs3Q8cUaXpKqAL+ApwCLAe+ExEHFJlHy0R0SNpJtmZyqcj4kFJvRHRXNbv5YgYcB6kvb09Ojs7B+pmZlYYRbgKS9KmimkEoI6X8UbEGQN0OQtA0jHAB2vsoyetX5J0J3Ai2dDXi5JmRcRuSbOAl0YucjOz4liyoLWwoxaNugprZlofAFwN3Fylz4GSDiqVyRLO9tS8HliaykuBu+ods5mZ7a9RcyAXS3oGeBroAb4O2ZCVpHtTn8OBhyQ9AfwM+GFE/Di13QCcKWkHcGZ6bWZmo6hucyBF5DkQM7OhqzUH4l+im5lZLk4gZmaWixOImZnl4gRiZma5OIGYmVkuTiBmZpaLE4iZmeXiBGJmZrk4gZiZWS5OIGZmlosTiJmZ5eIEYmZmuTiBmJlZLk4gZmaWixOImZnl4gRiZma5NOqRtsdLeljSNkl3S5pWpU+bpC1ly6uSrkhtn5fUXda2aNQ/hJnZBNeoM5BbgRURMQ+4E1he2SEiuiJifkTMB04Afpf6lnyp1B4R91Zub2Zm9dWoBNIGPJjKG4ALBuh/OvBsRPyirlGZmdmgNSqBbAfOTeULgdkD9L8I+HZF3eWStkq6XdL0WhtKWiapU1Lnnj178kdsZmb7UUTUZ8fS/cARVZquArqArwCHAOuBz0TEITX2MxXoAd4VES+musOBXwIBXAfMiohLB4qpvb09Ojs7c3waM7OJS9KmiGivrJ9crzeMiDMG6HIWgKRjgA/20+8c4PFS8kj7fqMs6R+Be4YRqpmZ5dCoq7BmpvUBwNXAzf10v5iK4StJs8penkc2JGZmZqOoUXMgF0t6BniabHjq6wCSWiS9cUWVpD8DzgTWVmy/Kl0CvBX4AHDl6IRtZmYldZsDKSLPgZiZDV2tORD/Et3MzHJxAjEzs1ycQMzMLJe6XcY7Xqzb3M3qji56evtoaW5i+cI2lixobXRYZmaDUs/vMCeQfqzb3M3Ktdvo27sPgO7ePlau3QbgJGJmhVfv7zAPYfVjdUfXG3/4kr69+1jd0dWgiMzMBq/e32FOIP3o6e0bUr2ZWZHU+zvMCaQfLc1NQ6o3MyuSen+HOYH0Y/nCNpqmTNqvrmnKJJYvbGtQRGZmg1fv7zBPovejNMnkq7DMbCyq93eYb2ViZmb98q1MzMxsRDmBmJlZLk4gZmaWixOImZnl4gRiZma5TKirsCTtAX4xCm91KPDLUXifvIocX5Fjg2LHV+TYoNjxFTk2aHx8fx4Rh1VWTqgEMlokdVa75K0oihxfkWODYsdX5Nig2PEVOTYobnwewjIzs1ycQMzMLBcnkPq4pdEBDKDI8RU5Nih2fEWODYodX5Fjg4LG5zkQMzPLxWcgZmaWixOImZnl4gQyAElnS+qStFPSiirtfyHpYUl/kPQ3FW3PS9omaYukzoq2T6f9PilpVZHik/TdVLcl9dlSoNjmS3qkVC/pxDyx1TG+49M22yTdLWlag+JrlvQDSU9LekrSe1P9DEkbJO1I6+kFiu3C9O/hj5KGdclqneJbneq2SrpTUnOBYrsuxbVF0n2SWvLENmQR4aXGAkwCngXeDkwFngCOregzE3gP8AXgbyrangcOrbLfDwD3A28p7aNI8VX0+SJwTVFiA+4DzknlRcADRfrbAY8Bp6TypcB1DYpvDfCxVJ4KNKfyKmBFKq8AbixQbO8E2oAHgPY8f7c6x3cWMDmVbyzY325aWZ/PADfn/fsNZfEZSP9OBHZGxHMR8RrwHWBxeYeIeCkiHgP2DmG/lwE3RMQfSvsoWHwASBLwIeDbBYotgNJR/cFAT47Y6hlfG/BgKm8ALhjt+NJZz/uB21K/1yKiNzUvJvsSIq2XFCW2iHgqIrpyxDNa8d0XEa+nro8ARxYotlfLuh5I9u+k7pxA+tcKvFD2eleqG6wA7pO0SdKysvpjgL+U9Kikf5H0noLFV/KXwIsRsaNAsV0BrJb0AnATsDJHbPWMbztwbipfCMxuQHxvB/YAX5e0WdKtkg5MbYdHxG6AtJ5ZoNhGymjEdynwoyLFJukL6d/Fh4FrcsQ2ZE4g/VOVuqFk9pMj4t3AOcCnJL0/1U8GpgMnAcuB76Wj/aLEV3Ix+c4+6hnbZcCVETEbuJJ0NFag+C5NrzcBBwGvNSC+ycC7ga9FxALg38mGq0ZKkWODOscn6SrgdeBbRYotIq5K/y6+BVyeI7YhcwLp3y72P4I8kiEMmURET1q/BNxJdvpa2u/ayPwM+CPZzdKKEh+SJgPnA9/NEVc9Y1sKrE3l75fHXIT4IuLpiDgrIk4gS77PNiC+XcCuiHg0vf4B2RcPwIuSZgGkdZ7h03rFNlLqFp+kpcB/Aj4cacKhKLGV+WfyD50OiRNI/x4D5ko6WtJU4CJg/WA2lHSgpINKZbIJuO2peR1wWmo7hmwyLM+dNusVH8AZwNMRsStHXPWMrQc4JZVPA/IMr9UtPkkz0/oA4Grg5tGOLyL+H/CCpLZUdTrw81ReT5aESeu7ChTbSKlLfJLOBv4bcG5E/K5gsc0t63ou8HTO+IZmNGbqx/JCdqXPM2RHklelur8G/jqVjyA7MngV6E3laWTjlU+k5cnStvHm1RPfJPvSeRw4rUjxpe3+qbSPIsUG/EdgU2p7FDihYPF9Nu3zGeAG0t0eRjO+1DYf6AS2kh2wTE/1hwAbyRLvRmBGgWI7L/X7A/Ai0FGwv91OsvmLLWnJdaVTnWK7g+z7ZCtwN9A6nH+7g118KxMzM8vFQ1hmZpaLE4iZmeXiBGJmZrk4gZiZWS5OIGZmlosTiJmZ5eIEYlaDpH1687b2W6rdensE36tZ0idHql+V7W6X9JKk7QP3Nhsc/w7ErAZJv42It43Se80B7omI40aiX5Xt3g/8FvjGULc1q2VyowMwm0jSrU++R3YPpElkzwv5Ltmv1t+h7OFdG4C5ZPdMeivw5Yi4Je2ist8TZM9/mEr2y/xPRsS+yveNiAdT8jEbMU4gZrU1af+nMV6fvuyH42ygJyI+CCDp4FS/AjguIuan+hkR8WtJTcBjku6IiF+V95P0TrIHRJ0cEXsl/S+yW3l/Y5gxmg2KE4hZbX2lL/RqJCmGPga8DbhJ0o1kQ1H/p0a/z0g6L5Vnk52R/Kqiz+nACWQJBqCJfHfXNcvFCcRsCCQdQXY7+fXAmvQwqRlkN737v2Q3yHsX8A7gK8ClEXFtafuIeEbSCWQ31Lte0n0R8T8q3uNUsrshvzcififpAbKhrD8JB1gTEXkfqmU2LL4Ky2xoFgDfiYgbyP79TCFLHicBL5PdrXcu2eNIPwbcUr6xpBbgdxHxTbInKpae5/AbsgdQQfao3pdT8viLtG+q9NsI/FXZLeRnSPrzkfuoZv3zGYhZbZVzID8GXiG7jTbAdWS3bz+MbJjpZWAZcDvZMxn2RXp8bJl5ZI/k/SNZkrkMICJ+Jelf02W2G4DJkrYCXWTP36ZKvx+RPXPkvvT8kb3Ap4BfVH4QSd8GTgUOlbQL+NuIyPs0RzPAl/GaDYmk24GPRcQfJX2O7IzjELIv+geAf4qIEyQ9BHw2IjY1Llqz+nICMTOzXDwHYmZmuTiBmJlZLk4gZmaWixOImZnl4gRiZma5OIGYmVkuTiBmZpaLE4iZmeXiBGJmZrn8f5pwJ/s81tlwAAAAAElFTkSuQmCC\n", "text/plain": ["
"]}, "metadata": {"needs_background": "light"}, "output_type": "display_data"}], "source": ["bhm.ll_E_scatter(bdata.models[2],1)"]}, {"cell_type": "raw", "id": "4f5a2588", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["By default the :func:`ll_E_scatter() ` plots only an evenly spaced distribution of parameter values, it does this by calling :meth:`ModelError.Loglik_Error.E_space() ` and plots only those values. You can pass the ``rng`` and ``steps`` keyword arguments to :func:`ll_E_scatter() ` , and these will be passed to :meth:`ModelError.Loglik_Error.E_space() ` , to adjust the values plotted.\n", "\n", "If on the other hand, you would like to see all values that have been evaluated, you can pass the keyword argument: ``rng_only=False`` , and it will plot all the values for the given parameter (everything stored in :attr:`ModelError.Loglik_Error.E_rng ` )"]}, {"cell_type": "code", "execution_count": 39, "id": "5d32a0d7", "metadata": {}, "outputs": [{"data": {"text/plain": [""]}, "execution_count": 39, "metadata": {}, "output_type": "execute_result"}, {"data": {"image/png": "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\n", "text/plain": ["
"]}, "metadata": {"needs_background": "light"}, "output_type": "display_data"}], "source": ["bhm.ll_E_scatter(bdata.models[2],1, rng_only=False)"]}, {"cell_type": "raw", "id": "46b2b68c", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Now let's see the same for the stoichiometry, same rules apply: "]}, {"cell_type": "code", "execution_count": 40, "id": "5cb2684c", "metadata": {}, "outputs": [{"data": {"text/plain": [""]}, "execution_count": 40, "metadata": {}, "output_type": "execute_result"}, {"data": {"image/png": "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\n", "text/plain": ["
"]}, "metadata": {"needs_background": "light"}, "output_type": "display_data"}], "source": ["bhm.ll_E_scatter(bdata.models[2],1, rng=3, steps=30)"]}, {"cell_type": "code", "execution_count": 41, "id": "ec9ddc4f", "metadata": {}, "outputs": [{"data": {"text/plain": [""]}, "execution_count": 41, "metadata": {}, "output_type": "execute_result"}, {"data": {"image/png": "iVBORw0KGgoAAAANSUhEUgAAAZAAAAEVCAYAAADOwrOnAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAei0lEQVR4nO3df5RcZZ3n8fcnnSZpUOmEH2IS2gAiKoRDpIUwjI4DSNAdIaCMKC6c9UfGX+s4akZywiissIhxRp3jOG7ccUYFNYom4CIG0EVnORJI7IQkYiSgQjoMIBAV0gMh+e4f9XSodKq6q2531b1V9XmdU6duP8/98e1bt+vb9z7Pfa4iAjMzs3pNyjsAMzNrTU4gZmaWiROImZll4gRiZmaZOIGYmVkmTiBmZpZJRyQQSedL2iRpt6T+KvNMlXSnpPVp3svL6j4p6W5J6yTdLGlGKn+dpLWSNqT308qWOTGVb5H0j5KUyqdIWp7KV0uaXbbMxZLuTa+Ly8qPSPPem5bdL5UrrXtLiu+VZcucJWlzqrukrHy6pFvSum6RNK2G/bcr/e7rJN1Q4243s3YXEW3/Al4OHAPcBvRXmUfA89J0N7AamJd+fkHZfB8EvpSm5wIz0vRxwGDZfHcCp6T13gS8PpW/r2z5C4DlaXo6cH96n5amp6W6bwMXpOkvAe9N029I6xYwD1idyruA+4Ajgf2A9cArUt2ngUvS9CXA1TXsvyfz/gz98suv4r064gwkIu6JiM1jzBMR8WT6sTu9ItX9oWzWA8rKByJiWyrfBExNZxgvopR0fhYRAXwNWJDmOwf4apq+Djg9nZ3MB26JiMcj4gngFuCsVHdampe0bPm6vpZivwPoTds+CdgSEfdHxDPAt9K8I7e/Z12SuiQtlXRXOpv5q9H2l5lZRySQWqUv0XXAI5S+zFeX1V0p6UHgQuDjFRZ/EzAQEU8DM4GtZXVbUxnp/UGAiHgW+D1wUHn5iGUOAraneauua0RdtXKAF0bEQ2n7DwGHpvJ3Ar+PiFcBrwLeLemIVDdV0hpJd0haUOF3N7MO1DYJRNKtkjZWeJ0z9tIlEbErIk4AZgEnSTqurG5JRBwOXAt8YMS2jwWuBob/a1el1Y9RV295lnWN5kzgopRAV1NKXEenur6I6AfeBnxO0lFjrMvMOsDkvAOYKBFxxgSua7uk24CzgI0jqr8B3Ah8AkDSLGAFcFFE3Jfm2UopCQ2bBWwrqzsc2CppMnAg8Hgqf+2IZW4Dfkfp0tTkdBZSaV0jt7NflXKAhyW9KCIeSpe7HknlAv57RKyqsD+2pff7036ZS6mNxcw6WNucgYyXpEMk9abpHuAM4Jfp56PLZj27rLyXUjJZHBG3D8+QLg39UdK81IZxEXB9qr4BGO5h9Wbgx6mdZBVwpqRpqWfUmcCqVPd/07ykZcvXdVHqjTWP0iWoh4C7gKNT7639KDXW31C2zMUV1rUKeK+k7vS7vVTSASmeKansYOBU4Be17lcza2N5t+I34wWcS+m/9aeBhyl9MQPMAH6Qpo8HBoC7KZ11fLxs+e+msruB7wMzU/mlwFPAurLXoamuPy1zH/AFQKl8KvAdYAulnlpHlm3nHal8C/DfysqPTPNuSctOSeUC/iltYwNlPcwo9dD6VapbUlZ+EPAj4N70Pj2VTwL+Z1rPRkpJ60DgT1LZ+vT+zrw/T7/88qsYr+EvNTMzs7r4EpaZmWXiBGJmZpm0RS+sgw8+OGbPnp13GGZmLWXt2rW/i4hDsi7fFglk9uzZrFmzJu8wzMxaiqTfjmd5X8IyM7NMnEDMzCwTJxAzM8vECcTMzDJpi0Z0s6K4dOUGvrn6QXZF0CXx1pMP54oFc6qWm7UyJxCzDColBIBr7nhgzzy7IrjmjgdYff9j3PvIU/uUD3NisVZVuKFMJJ1A6al7U4FngfdFxJ2jLdPf3x/uxmuNUEuiGCbGHjO/Fm+f1wc4sVjjSVobpUc1ZFu+gAnkZuCzEXGTpDcAfxsRrx1tGScQmygrBwZZumoz27YPMbV7EkM7d+8zz0QlinodsF8XO57ZxYzeHhbNP4YFc2eOvZDZKMabQIp4CSuAF6TpA3nuORZmDbFyYJDLbtjE9qGde5VXSh6QT/IAeOqZXQAMbh/iQ8vX8aHl65i2fzefeOOxTiaWiyImkA8BqyR9hlIvsT+pNJOkhcBCgL6+vqYFZ+1l5cAgi76znp27G5cWjj70gL3aQIZNxJnMEzt2sui69QBOItZ0uVzCknQrcFiFqiXA6cBPIuK7kv4SWBhjPG3Ql7CsVuWXqGb09rDjmWd5YsfOsReswWhtF41uS5nZ28Ptl5w2zt/AOk07toH8HuiNiEhP8/t9RLxgtGWcQGwsl67cUPHLOouJbOSuJ7GMpWdEm80B+3Vx5blzfGZiVbVjArkHeG9E3CbpdODTEXHiaMs4gVg1pUtU66jSnDGmSQCC3UFTe0OVJxYBkyaJXRkvs7mdxKppxwTyp8DnKbXP/CelbrxrR1vGCcRGqtYwXo8iffGO9/fp6e7iqvN8NmJ7a7sEkoUTiJW78Ms/4/b7Hq95/t6eboA9X85FShwjjUwk0/bvrqsNx5e1rFw7duM1y+zSlRvqSh4CLju7mMmikgVzZ+4T66mf+jGD24dqWv6pZ3bxke+415ZNDJ+BWMsr71lV79H89nl9LX+H98qBQT68fB31NPN0SeyO8E2JHc6XsHAC6WSXrtzAtXc8UHfimDJ5Ele/6fi2+eLMkkTKFfmynTWOL2FZR1o5MMji791d9W7xatr1i3L40lZ57616PLFjJx9evm7Pusxq4TMQazlZ7h4/9ajpXPvuUxoYVfFkvfdlpi9rdQyfgVhHWTkwyEe+vb7m/7A7eSTb4d+5/BJfLXe5D24fYvH3NgA+G7HR+QzEWkI990EI+OxbTvCXXwWlS38bGNq5a8x5PTxK+xvvGYgfaWuFd+nKDfzN8nU130R34bw+J48qFsydyVXn1XY2Nrh9iKMW/4BLV25ocFTWqpxArLBWDgxywuU3c00dvazaoVtuoy2YO3PPeF5jGX564oVf/lmDo7JW5ARihTR8qaXWs45p+3fzubec4ORRoysWzOHUo6bXPP/t9z3OyoHBBkZkrcgJxApp6arNNV2n7+nu4nNvOYGBj5/py1Z1uvbdp/C5t5zAzN4eoNThYDRLV21uRljWQtwLywpl+K7yWobmaNd7Oppp5NAoRy3+QdUebttqHC7FOocTiBVGrXeVi1JDuS9XTby3nnx41XtHpnZP2pNgOrl7tD3HCcQKYeXAYE3Jw2cdjXXFgjn8+tEn9xmQchJ7PyN+uHF9eBnrTG4DsUJYumrzqMljZm+P2zqapLxtRJT2PVWaR765+sGmxmbF4jMQK4TRrq/7hrbmG9k2MvuSGyvOV++YW9ZenECsEGb09lRsOBewaP4xzQ/I9tIlVUwWovQ8km3bhzw0fAdyArFclI8a2yUx78hpPP7UM3t13R1uLPcXUv6qNa5PmqQ9iX9w+xB/s3wda377uNtFOoQTiDVVpWHYd0Vw+32Pc+pR0/nNY0P+b7aAhhNCedKf2j2Jp57Z+16dADeud5BcBlOUdD5wGfBy4KSIWFNWtxh4J7AL+GBErBprfR5MsTWMNZBfl8R9V72hyVFZVkdccmPVjg8e0LI1tOpgihuB84CflhdKegVwAXAscBbwRUldzQ/PGmGsu8vdINtaZqQ72CsJfOd6J8glgUTEPRFR6eg6B/hWRDwdEb8GtgAnNTc6a5Sx7mQeaygNK5ZF84+p1rsX8J3rnaBo94HMBMo7lm9NZdYGRvuPFUoNtdY6FsydyYWjjOo71udtra9hjeiSbgUOq1C1JCKur7ZYhbKK1zUkLQQWAvT11TY0tTXf8NhW27YPcWBPN91dYueuvT/SSYK3neyhSVpRpaceQmmQyz9/2SHu4tvmGpZAIuKMDIttBcr/DZ0FbKuy/mXAMig1omfYljXYyEbz7UM76Z4kpu3fzfYdO/2l0iauWDCH/hdP3/OPwozeHv78ZYfw3bWDez57Pya3PRWtG+8NwDck/QMwAzgauDPfkCyrSo3mO3cH++83mYGPn5lTVNYII+9cP/VTP97nsx/auYulqzY7gbSRXNpAJJ0raStwCnCjpFUAEbEJ+DbwC+CHwPsjYuyHQlghVWtEdeNq+/Nn3xny6oW1IiJmRcSUiHhhRMwvq7syIo6KiGMi4qY84rOJUa0R1Y2r7c+ffWcoWi8sayOL5h9DT/fet/H0dHd5bKsO4M++MxStDcTayPC17vLGVTeadwZ/9p0hl6FMJpqHMslfeXddf1nYaHysFMd4hzLxGYiN28juuu6yadX4WGkvbgOxcavUXXe4y6ZZOR8r7cUJxMbNXTatVj5W2osTiI2bu2xarXystBcnEBs3d9m0WvlYaS9uRLdxc5dNq5WPlfbibrxWM3e/tEa6dOWGvR6Z+9aTD/cIzQ3mbrzWFCsHBll03fo9Q7EPbh9i0XXrAXe/tPG7dOWGPc9Sh9LTKf1s9eJzG4jV5PLvb9rnOR47dwWXf39TThFZO/nm6gfrKrdicAKxmjyxY2dd5Wb12FXlUnq1cisGJxAzy12XKj9dvVq5FYMTiNWkt6e7rnKzerz15MPrKrdicAKxmlx29rF0T9r7v8HuSeKys4/NKSJrJ1csmMPb5/XtOePoknj7vD43oBecu/FazdyN16y9uBuvTbhqiWLkc6/NmsH/uBSXE4jtxcNtW5H4eCw2t4HYXjzcthWJj8diyyWBSDpf0iZJuyX1l5W/TtJaSRvS+2l5xNfJPNy2FYmPx2LL6wxkI3Ae8NMR5b8D3hgRc4CLga83O7BO5+G2rUh8PBZbLgkkIu6JiH3OQSNiICK2pR83AVMlTWludJ3Nw21bkfh4LLYiN6K/CRiIiKcrVUpaCCwE6Ovra2Zcbc3DbVuR+HgstobdByLpVuCwClVLIuL6NM9twEcjYs2IZY8FbgDOjIj7xtqW7wMxM6tfYe8DiYgzsiwnaRawArioluRhZp3J94fkr1CXsCT1AjcCiyPi9pzDaXv+A7RW5ftDiiGvbrznStoKnALcKGlVqvoA8BLg7yStS69D84ix3Q3/AQ5uHyJ47g9w5cBg3qGZjcn3hxRDXr2wVkTErIiYEhEvjIj5qfyKiDggIk4oez2SR4ztzn+A1sp8f0gx+E70DuU/QGtlvj+kGJxAOpT/AK2V+f6QYnAC6VD+A7RWtmDuTK46bw4ze3sQMLO3h6vOm+MG9CYrVC8sax7foGWtzo8XyJ8TSAfzH6CZjYcvYZmZWSZOIGZmlokTiJmZZeIEYmZmmTiBmJlZJk4gZmaWiROImZll4gRiZmaZOIGYmVkmTiBmZpaJE4iZmWXiBGJmZpk4gZiZWSZOIGZmlkkuCUTS+ZI2Sdotqb9CfZ+kJyV9NI/4zMxsbHmdgWwEzgN+WqX+s8BNzQvHzMzqlcsDpSLiHgBJ+9RJWgDcDzzV3KjMzKwehWoDkXQA8DHg8rxjMTOz0TXsDETSrcBhFaqWRMT1VRa7HPhsRDxZ6exkxPoXAgsB+vr6xhOqmZll0LAEEhFnZFjsZODNkj4N9AK7Jf1nRHyhwvqXAcsA+vv7YzyxmplZ/XJpA6kmIl49PC3pMuDJSsnDzMzyl1c33nMlbQVOAW6UtCqPOMzMLLu8emGtAFaMMc9lzYnGzMyyKFQvLDMzax1OIGZmlokTiJmZZZI5gUh6YCIDMTOz1jKeM5DR7/QzM7O2Np4E4pv3zMw62KjdeCV9uFoV8LyJD8fMzFrFWPeBPH+Uus9PZCBmZtZaRk0gEVF1VFxJH5rwaMzMrGWMpw2k2uUtMzPrAO6FZWZmmbgXlpmZZTJWL6w/UjlRCOhpSERmZtYSxmpEH60XlpmZdTCPhWVmZpk4gZiZWSZOIGZmlokTiJmZZeIEYmZmmTiBmJlZJrkkEEnnS9okabek/hF1x0v6WarfIGlqHjGamdnoxhqNt1E2AucB/6u8UNJk4Brgv0bEekkHATtziM/MzMaQSwKJiHsApH2G0zoTuDsi1qf5HmtyaGZmVqOitYG8FAhJqyT9XNLfVptR0kJJayStefTRR5sYopmZQQPPQCTdChxWoWpJRFw/Sjx/CrwK2AH8SNLaiPjRyBkjYhmwDKC/v98DO5qZNVnDEkhEnJFhsa3ATyLidwCSfgC8EtgngZiZWb6KdglrFXC8pP1Tg/qfAb/IOSYzM6sgr26850raCpwC3ChpFUBEPAH8A3AXsA74eUTcmEeMZmY2urx6Ya0AVlSpu4ZSV14zMyuwol3CMjOzFuEEYmZmmTiBmJlZJk4gZmaWiROImZll4gRiZmaZOIGYmVkmTiBmZpaJE4iZmWXiBGJmZpk4gZiZWSZOIGZmlokTiJmZZeIEYmZmmTiBmJlZJk4gZmaWiROImZll4gRiZmaZOIGYmVkmuSQQSedL2iRpt6T+svJuSV+VtEHSPZIW5xGfmZmNLa8zkI3AecBPR5SfD0yJiDnAicBfSZrd5NjMzKwGk/PYaETcAyBpnyrgAEmTgR7gGeAPzY3OzMxqUbQ2kOuAp4CHgAeAz0TE45VmlLRQ0hpJax599NFmxmhmZjTwDETSrcBhFaqWRMT1VRY7CdgFzACmAf8u6daIuH/kjBGxDFgG0N/fHxMTtZmZ1aphCSQizsiw2NuAH0bETuARSbcD/cA+CcTMzPJVtEtYDwCnqeQAYB7wy5xjMjOzCvLqxnuupK3AKcCNklalqn8Cnkepl9ZdwL9GxN15xGhmZqPLqxfWCmBFhfInKXXlNTOzgivaJSwzM2sRTiBmZpaJE4iZmWXiBGJmZpk4gZiZWSZOIGZmlokTiJmZZeIEYmZmmTiBmJlZJk4gZmaWiROImZll4gRiZmaZOIGYmVkmTiBmZpaJE4iZmWXiBGJmZpk4gZiZWSZOIGZmlokTiJmZZZJLApG0VNIvJd0taYWk3rK6xZK2SNosaX4e8ZmZ2djyOgO5BTguIo4HfgUsBpD0CuAC4FjgLOCLkrpyitHMzEaRSwKJiJsj4tn04x3ArDR9DvCtiHg6In4NbAFOyiNGMzMbXRHaQN4B3JSmZwIPltVtTWVmZlYwkxu1Ykm3AodVqFoSEdeneZYAzwLXDi9WYf6osv6FwEKAvr6+ccdrZmb1aVgCiYgzRquXdDHwF8DpETGcJLYCh5fNNgvYVmX9y4BlAP39/RWTjJmZNU5evbDOAj4GnB0RO8qqbgAukDRF0hHA0cCdecRoZmaja9gZyBi+AEwBbpEEcEdEvCciNkn6NvALSpe23h8Ru3KK0czMRpFLAomIl4xSdyVwZRPDMTOzDIrQC8vMzFqQE4iZmWXiBGJmZpk4gZiZWSZOIGZmlokTiJmZZeIEYmZmmeR1I6GZmY3DyoFBlq7azLbtQ8zo7WHR/GNYMLe5Y886gZiZtZiVA4Ms/t4GhnaWBuoY3D7E4u9tAGhqEvElLDOzFrN01eY9yWPY0M5dLF21ualxOIGYmbWYbduH6ipvFCcQM7MWM6O3p67yRnECMTNrMYvmH0NPd9deZT3dXSyaf0xT43AjuplZixluKHcvLDMzq9uCuTObnjBG8iUsMzPLxAnEzMwy6ehLWEW4k9PMrFV1bAIpyp2cZmatqmMvYRXlTk4zs1aVSwKRtFTSLyXdLWmFpN5U/jpJayVtSO+nNSqGotzJaWbWqvI6A7kFOC4ijgd+BSxO5b8D3hgRc4CLga83KoCi3MlpZtaqckkgEXFzRDybfrwDmJXKByJiWyrfBEyVNKURMRTlTk4zs1ZVhEb0dwDLK5S/CRiIiKcrLSRpIbAQoK+vr+6NFuVOTjOzVqWIaMyKpVuBwypULYmI69M8S4B+4LwoC0TSscANwJkRcd9Y2+rv7481a9ZMTOBmZh1C0tqI6M+6fMPOQCLijNHqJV0M/AVw+ojkMQtYAVxUS/IwM7N85HIJS9JZwMeAP4uIHWXlvcCNwOKIuD2P2MzMrDZ59cL6AvB84BZJ6yR9KZV/AHgJ8HepfJ2kQ3OK0czMRpHLGUhEvKRK+RXAFU0Ox8zMMujYO9HNzGx8GtYLq5kkPQr8tgGrPpjSzY1F47jqV9TYHFd9HFd9xorrxRFxSNaVt0UCaRRJa8bTxa1RHFf9ihqb46qP46pPo+PyJSwzM8vECcTMzDJxAhndsrwDqMJx1a+osTmu+jiu+jQ0LreBmJlZJj4DMTOzTJxAzMwsk7ZOIJLOkrRZ0hZJl1Sof5mkn0l6WtJHR9T9Jj0ZcZ2kNWXl0yXdIune9D6trG5x2tZmSfObHFe1pzzOljRUNjTMlxhFg2K7TNJgWQxvKMg+W14W028krat3n40zrl5J16XP7R5Jp6TyvI+xanGN+xhrUFx5H1/V4srt+JJ0TNn610n6g6QPpbpxH197RERbvoAu4D7gSGA/YD3wihHzHAq8CrgS+OiIut8AB1dY76eBS9L0JcDVafoVaRtTgCPStruaGNeZwOQ0fXVZXLOBjTnvs8tGzluEfTZinr8HPl7PPpuAuL4KvCtN7wf0FuQYqxbXuI6xBsaV9/FVMa68j68R6/kPSjcNjvv4Kn+18xnIScCWiLg/Ip4BvgWcUz5DRDwSEXcBO+tY7zmUDhjS+4Ky8m9FxNMR8WtgS4qhKXFFlac81qlR+6yaXPfZMEkC/hL4Zp2LZo5L0guA1wD/kuZ7JiK2p+rcjrHR4pqAY6xR+6uaXPdX2TxNP75GOB24LyKGR+sY7/G1RzsnkJnAg2U/b01ltQrgZklrVXr64bAXRsRDAOl9eLTgWrfXqLjKvQO4qeznIyQNSPqJpFePsu5GxvaBdOnjK2WnzEXZZ68GHo6Ie8vKatln44nrSOBR4F/Tdv63pANSXZ7H2GhxlctyjDUyrryOr1r2Vx7HV7kL2Dt5jff42qOdE4gqlNXTZ/nUiHgl8Hrg/ZJeM0Hba2hcKj3l8Vng2lT0ENAXEXOBDwPfSP81VdKo2P4ZOAo4IcXz93Vur9Gf5VvZ+w+s1n02nrgmA68E/jlt5ylKlxNG04z9NWZc4zjGGhVXnsdXLZ9jHsdXaQXSfsDZwHdqmb3e7bVzAtkKHF728yxgW60LR8S29P4IpSckDp/KPSzpRQDp/ZE6t9eouMqf8nhhpIua6XT0sTS9ltJ1zZdWWX1DYouIhyNiV0TsBr5cFnMR9tlk4Dxgedn8te6z8cS1FdgaEavTz9dR+iKCfI+x0eIa7zHWkLhyPr7G2l95HV/DXg/8PCIeLisb7/G1RzsnkLuAoyUdkbLwBZSesz4mSQdIev7wNKXGw42p+gbg4jR9MXB9WfkFkqZIOgI4GrizWXHpuac8nh17P+XxEEldafrIFNf9VTbRqNheVDbruey9L3PbZ8kZwC8jYmvZMrXus8xxRcR/AA9KOiYVnQ78Ik3ndoyNFtcEHGONiiu342uMzxFyOr7KjDz7gfEfX8+p1rreDi/gDcCvKGX4JansPcB70vRhlLLuH4DtafoFlK5rrk+vTcPLpmUOAn4E3Jvep5fVLUnb2gy8vslxbaF0/XJden0plb8pzbse+Dnwxhz22deBDcDd6SB9URH2WVru34bXUVZW8z7LGleqOwFYk/bLSmBa3sfYGHGN+xhrUFy5HV+jxVWA42t/4DHgwBHrHPfxNfzyUCZmZpZJO1/CMjOzBnICMTOzTJxAzMwsEycQMzPLxAnEzMwycQIxM7NMnECs40japb2Huh5rCJHxbKtX0vsmar4Ky31F0iOSNo49t9nE8n0g1nEkPRkRz2vStmYD/ycijpuI+Sos9xrgSeBr9S5rNl6T8w7ArB2kYVK+TWn8oC7gkxGxHPgUcJRKDxO6hdLwEIcDU4HPR8SytIqR860HPkjpORCrgfdFxK6R242In6bkY9Z0TiDWiXrSF/Wwq9KX/XicBWyLiP8CIOnAVH4JcFxEnJDKp0fE45J6gLskfTdKA+vtmU/Syyk99OfUiNgp6YvAhcDXxhmj2YRyArFONDT8hV6JJEX913Y3AJ+RdDWlS1H/XmW+D0o6N00fTumM5LER85wOnEgpwQD08NyIqWaF4QRiBkg6DPgepcH4vqrSg6emUxqg7teUBpg7ltJzJ/4ReEdEXD68fET8StKJlAa/u0rSzRHxP0Zs47WURmc9JSJ2SLqN0qWsfcIBvhoRiyfydzSbaO6FZVYyl9LjPD9F6e+im1LymAc8QWlk36MpPTr0XcCy8oUlzQB2RMQ1wGd47pkQfwSen6YPBJ5IyeNlad1UmO9HwJslHZrWPV3SiyfuVzWbGD4DsU40sg3kh8DvKQ3FDfBJ4K+BQyhdZnoCWAh8hdLT3XZFeiRomTnAUkm7KSWZ9wJExGOSbk/dbG8BJku6m9Jw2XcMLzxivpuASyk9hndSWt/7gd8ygqRvAq8FDpa0FfhERPxLlp1iVi934zWjdD8F8K6I2C3pI5TOOA6i9EV/G/BvEXGipP8H/HWUniRn1tGcQMzMLBO3gZiZWSZOIGZmlokTiJmZZeIEYmZmmTiBmJlZJk4gZmaWiROImZll4gRiZmaZOIGYmVkm/x8f6UMn96MJfAAAAABJRU5ErkJggg==\n", "text/plain": ["
"]}, "metadata": {"needs_background": "light"}, "output_type": "display_data"}], "source": ["bhm.ll_E_scatter(bdata.models[2],1, rng=3, steps=30, rng_only=False)"]}, {"cell_type": "raw", "id": "e4e57357", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["And for transition rates, note that :func:`ll_trans_scatter() ` needs two indices for ``from_state`` and ``to_state`` , instead of just one."]}, {"cell_type": "code", "execution_count": 42, "id": "b90f06d4", "metadata": {}, "outputs": [{"data": {"text/plain": [""]}, "execution_count": 42, "metadata": {}, "output_type": "execute_result"}, {"data": {"image/png": "iVBORw0KGgoAAAANSUhEUgAAAaAAAAEPCAYAAAAEfBBiAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAhWklEQVR4nO3df5RV5b3f8ffnMgJTsRkVmvAjgrEsatQIzRGu1buupSyHeldhIOBt06S4Lrksq6nr1i6WspK1TJOsqqG9pjdNiiS6Qm9zqzdcRRKvTjRqTXMrcQggIFKNNRFoKwQmGhkSfnz7x35GNzPnnDkzc87sw/B5rbXXnP3d+3n285z58Z29n+fsrYjAzMxspP1O0Q0wM7OzkxOQmZkVwgnIzMwK4QRkZmaFcAIyM7NCOAGZmVkhWopuwJlk4sSJMWPGjKKbYWZ2Rtm6deuhiJjUN15IApK0HPgCcCkwNyK6UnwusL53N+ALEfFo2vYkMJmszT8Cbo2Ik5JuBz4DnAAOAn8UET9PZVYAn0/1fTkiNqT4xcBDwAXAT4FPR8RvB2r3jBkz6OrqGmbvzczOLpJ+Xi5e1CW4XcBS4Pky8VJEzAYWAvdL6k2SN0bElcDlwCRgeYpvS2U+BmwEvgIg6QLgLmAeMBe4S9L5qcy9wH0RMRM4Aqysew/NzKyqQhJQROyJiL1l4kcj4kRaHQ9Ebtvb6WULMLZ3W0Q8GxFH07YXgGnpdTvwVEQcjogjwFPAQkkC5pMlK4ANQEe9+mZmZrVpukkIkuZJ2g3sBG7OJSQkdQJvAe/wfgLJWwk8kV5PBd7MbduXYhcC3bl6e+NmZjaCGpaAJD0taVeZZXG1chGxJSIuA64C1kgan9vWTjYONI7sLCZ/vE8BJWBtb6hc9VXilfqxSlKXpK6DBw9Wa7qZmQ1CwyYhRMSCYZbfI+ldsjGfrlz8mKTNwGKyy2pIWgB8Dvj9iPhN2nUfcF2uymnAc8AhoE1SSzoLmgYcqNKO9aSJEaVSyXduNTOrk6aahp1mp70ZESckTQdmAW9ImgCcFxH/J01KuIFsJhyS5gD3Awsj4q1cdZ3Av8tNPLgeWBMRIelZYBnZTLgVwGMj0T+zwdq0bT9rO/dyoLuHKW2trG6fRcec2q8YD7e8WSMVNQ17CfA1stlsj0vani6vXQvcKek4cAq4JSIOSfogsFnSOGAM8AywLlW3FpgAfDebX8AvImJRRByW9CXgxbTfFyPicHp9B/CQpC+TzaJ7oNF9NhusTdv2s+aRnfQcPwnA/u4e1jyyE6CmJDLc8maNJj8PqHalUin8OSAbjOGcgVxzzzPs7+7pF5/a1sqP75xfpkR9y4PPoKw+JG2NiFLfeFNdgjMbTYZ7BnKgTPKoFq93eZ9BWaM13TRss9Fibefe9/549+o5fpK1nf0+AlfWlLbWQcXrXX647TcbiBOQWYMM9wxkdfssWs8Zc1qs9ZwxrG6fNSLlh9t+s4H4EpxZFcMZA5nS1lp2DKbWM5De4wz1+MMtP9z2mw3EkxAGwZMQzi59x0AgO4O4e+kVQ5qFNtjyRTvT22/Nw5MQzAap2hhILX+Ah3sGUrR6tN+z6KwaJyCzCuoxBtIxZ+oZ/Qd3OO33LDobiCchmFUw3FlkZzvPorOBOAGZVTDcWWRnO8+is4H4EpyNasMZgzjTx3CK5ll0NhAnIBu16jEGcaaP4RRpdfussrPofAZpvXwJzkYtj0EUq2POVO5eegVT21oR2T3oPIXb8nwGZKOWxyCK5zNIq8ZnQDZqeRabWXNzArJRy7PYzJqbL8HZqOVZbGc+30lhdHMCslHNYxBnLt9JYfRzArKm5v+Az17DvRefNT8nIGta/g/47OZZjKNfIZMQJC2XtFvSKUmlXHyupO1p2SFpSW7bkym2W9I6SWNS/HZJL0t6SdIPJU3PlTmZq29zLn6xpC2SXpX0sKSxI9V3q50/x3N28yzG0a+oWXC7gKXA82XipYiYDSwE7pfUe5Z2Y0RcCVwOTAKWp/i2VOZjwEbgK7n6eiJidloW5eL3AvdFxEzgCLCyfl2zevF/wGc3z2Ic/QpJQBGxJyL6/RsbEUcj4kRaHQ9Ebtvb6WULMLZ3W0Q8GxFH07YXgGnVji1JwHyyZAWwAegYWk+skfwf8NnNd1IY/ZpuDEjSPOBBYDrw6VxCQlInMBd4gvcTSN7KtK3XeEldwAngnojYBFwIdOfq3Qf4J7oJ+V5i5lmMo1vDEpCkp4EPldn0uYh4rFK5iNgCXCbpUmCDpCci4lja1i5pPPAdsrOYp3LH+xRQAn4/V91FEXFA0keAZyTtBN6mv4rPJZe0ClgFcNFFF1XazRrAn+MxG90aloAiYsEwy++R9C7ZmE9XLn4sTShYTEpAkhYAnwN+PyJ+k9v3QPr6uqTngDnAXwFtklrSWdA04ECVdqwH1gOUSqWKicoaw/8Bm41eTXUrnjQ7rSW9ng7MAt6QNEHS5BRvAW4AXknrc4D7gUUR8VaurvMljUuvJwLXAC9HRADPAsvSriuAimdkNjybtu3nmnue4eI7H+eae55h07b9RTfJzJpEUdOwl0jaB1wNPJ7GdgCuBXZI2g48CtwSEYeAc4HNkl4CdgBvAetSmbXABOC7faZbXwp0SdpBlnDuiYiX07Y7gNslvUY2JvRAA7t71ur9HM/+7h6C9z/H4yRkZgDKTgisFqVSKbq6ugbe0QC45p5nyj4Rc2pbKz++c34BLTKzIkjaGhGlvvGmugRno4s/x2Nm1TgBWcP4czxmVo0TkDWMP8luRfMkmObWdB9EtdHDn+OxIvlmts3PCcgayp/jsaL4cQ7Nz5fgzGxU8iSY5ucEZGajkifBND8nIKvKg7h2pvIkmObnMSCryIO4dibzJJjm5wRkFXkQ1850ngTT3HwJziryIK6ZNZITkFXkQVwzayQnIKvIg7hm1kgeA7KKPIhrZo3kBGRVeRDXzBrFl+DMzKwQTkBmZlYIX4Ib5TZt2+8xHDNrSk5Ao5jvZGBmzayQS3CSlkvaLemUpFIuPlfS9rTskLQkt+3JFNstaZ2kMSl+s6Sdqcz/kPTRXJkVkl5Ny4pc/GJJW1L8YUljR6rvI6nanQzMzIpW1BjQLmAp8HyZeCkiZgMLgfsl9Z6l3RgRVwKXA5OA5Sn+FxFxRSrzFeBPASRdANwFzAPmAndJOj+VuRe4LyJmAkeAlXXvYRPwnQzMrJkVkoAiYk9E9Ps3PCKORsSJtDoeiNy2t9PLFmBs77ZcHODcXJl24KmIOBwRR4CngIWSBMwHNqb9NgAd9ehXs/GdDMyGx3eDb6ymmwUnaZ6k3cBO4OZcQkJSJ/AW8A7vJxAk3SrpZ2RnQLel8FTgzVzV+1LsQqA7V29vvFJ7VknqktR18ODBYfdvJPlOBmZD1zuGur+7h+D9MVQnofppWAKS9LSkXWWWxdXKRcSWiLgMuApYI2l8bls7MBkYR3YW0xv/ekRcAtwBfL63CeWqrxKv1J71EVGKiNKkSZOqNb3pdMyZyt1Lr2BqWysCpra1cvfSKzwBwawGHkNtvIbNgouIBcMsv0fSu2RjPl25+DFJm4HFZJfV8h4C/nN6vQ+4LrdtGvAccAhok9SSzoKmAQeG09Zm5jsZmA2Nx1Abr6kuwaXZaS3p9XRgFvCGpAmSJqd4C3AD8Epan5mr4g+AV9PrTuB6SeenyQfXA50REcCzwLK03wrgscb2zMzONB5DbbyipmEvkbQPuBp4PI3tAFwL7JC0HXgUuCUiDpFNLtgs6SVgB9k40LpU5rNpavZ24HayhEJEHAa+BLyYli+mGGSX6m6X9BrZmNADjeyvmZ15PIbaeMpOCKwWpVIpurq6Bt6xjnwnA7Pi+PevPiRtjYhS37jvhNDEfCcDs2J5DLWxmmoMyE7nWThmNpo5ATUxz8Ixs9HMCaiJeRaOmY1mTkBNzLNwzGw08ySEJtY7+OlZOGY2GjkBNTnPwjGz0cqX4MzMrBBOQGZmVggnIDMzK4QTkJmZFcKTEBrM95IyMyvPCaiBfC83M7PKfAmugXwvNzOzypyAGsj3cjMzq8wJqIF8Lzczs8qcgBrI93IzM6vMkxAayPdyMzOrrJAEJGk58AXgUmBuRHSl+Fxgfe9uwBci4tG07UlgMlmbfwTcGhEnJd0M3AqcBH4NrIqIl1OZk8DOVN8vImJRil8MPARcAPwU+HRE/LYRffW93MzOXv4YRnVFXYLbBSwFni8TL0XEbGAhcL+k3iR5Y0RcCVwOTAKWp/hfRMQVqcxXgD/N1dcTEbPTsigXvxe4LyJmAkeAlfXrmpnZ+x/D2N/dQ/D+xzA2bdtfdNOaRiEJKCL2RES/ucgRcTQiTqTV8UDktr2dXrYAY3u35eIA5+bLlCNJwHxgYwptADoG3wszs8r8MYyBNd0kBEnzJO0mu3R2cy4hIakTeAt4h/cTCJJulfQzsjOg23LVjZfUJekFSR0pdiHQnat3H+BzYjOrK38MY2ANS0CSnpa0q8yyuFq5iNgSEZcBVwFrJI3PbWsnGwcaR3YW0xv/ekRcAtwBfD5X3UURUQI+CXxV0iVkY0v9DlulH6tSEus6ePDgwB03M8Mfw6hFwxJQRCyIiMvLLI/VWH4P8C7ZmE8+fgzYDJRLZA+Ru5wWEQfS19eB54A5wCGgLTe2NA04UKUd6yOiFBGlSZMm1dJ0MzN/DKMGTXUJTtLFvYlB0nRgFvCGpAmSJqd4C3AD8Epan5mr4g+AV1P8fEnj0uuJwDXAyxERwLPAslRmBVBTUjQzq1XHnKncvfQKpra1ImBqWyt3L73Cs+ByipqGvQT4GtlstsclbU+X164F7pR0HDgF3BIRhyR9ENicEsoY4BlgXarus5IWAMfJZrStSPFLyWbRnSJLtPf0Ts8mu1T3kKQvA9uABxrcZTM7C/ljGNUpOyGwWpRKpejq6iq6GWZmZxRJW9N4/Gma6hKcmZmdPZyAzMysEE5AZmZWCCcgMzMrhBOQmZkVwgnIzMwK4QRkZmaFcAIyM7NCOAGZmVkhnIDMzKwQTkBmZlYIJyAzMyvEkBOQpF/UsyFmZnZ2Gc4ZULkni5qZmdVkOAnIz3EwM7Mhq/pAOkm3V9oETKh/c8zM7Gwx0BNRz6uy7T/WsyFmZnZ2qZqAIuLfVtom6U/q3hozMztrDGcMqNLlOTMzswEVMgtO0nJJuyWdklTKxedK2p6WHZKW5LY9mWK7Ja2TNKZPncskRZ/6Vkh6NS0rcvGLJW1J8YcljR1qX8zMbGiKmgW3C1gKPF8mXoqI2cBC4H5JvZcJb4yIK4HLgUnA8t5Cks4DbgO25GIXAHcB84C5wF2Szk+b7wXui4iZwBFg5TD6YmZmQ1A1AUl6R9LbZZZ3gClDPWhE7ImIvWXiRyPiRFodTy7JRcTb6WULMJbTE+CXgK8Ax3KxduCpiDgcEUeAp4CFkgTMBzam/TYAHUPti5mZDU3VBBQR50XE3y6znBcRA82gGxJJ8yTtBnYCN+cSEpI6gbeAd0gJRNIc4MMR8f0+VU0F3syt70uxC4HuXL29cTMzG0ENSSIAkp4GPlRm0+ci4rFK5SJiC3CZpEuBDZKeiIhjaVu7pPHAd4D5kn4I3AfcVK4J5aqvEq/Uj1XAKoCLLrqo0m5mZnW3adt+1nbu5UB3D1PaWlndPouOOaPn/+WGJaCIWDDM8nskvUs25tOVix+TtBlYDPwkbX8uu7LGh4DNkhaRndlcl6tyGvAccAhok9SSzoKmAQeqtGM9sB6gVCr57g9mNiI2bdvPmkd20nP8JAD7u3tY88hOgFGThJrqbthpdlpLej0dmAW8IWmCpMkp3gLcALwSEb+KiIkRMSMiZgAvAIsiogvoBK6XdH6afHA90BkRATwLLEuHXQFUPCMzMyvC2s697yWfXj3HT7K2s9/w+RmrkAQkaYmkfcDVwONpbAfgWmCHpO3Ao8AtEXEIOJfszOYlYAfZONC6aseIiMNkkxNeTMsXUwzgDuB2Sa+RjQk9UM/+mZkN14HunkHFz0QNuwRXTUQ8SpZg+sb/HPjzMvH/B1xVQ73X9Vl/EHiwzH6vk03NNjNrSlPaWtlfJtlMaWstoDWN0VSX4MzMLLO6fRat55z2eXtazxnD6vZZBbWo/go5AzIzs+p6Jxp4FpyZmY24jjlTR1XC6cuX4MzMrBBOQGZmVggnIDMzK4QTkJmZFcIJyMzMCuEEZGZmhXACMjOzQjgBmZlZIZyAzMysEE5AZmZWCCcgMzMrhBOQmZkVwgnIzMwK4QRkZmaFcAIyM7NCOAGZmVkhCklAkpZL2i3plKRSLj5X0va07JC0JLftyRTbLWmdpDF96lwmKfrUdzJX3+Zc/GJJWyS9KulhSWMb3WczMztdUWdAu4ClwPNl4qWImA0sBO6X1PvU1hsj4krgcmASsLy3kKTzgNuALX3q64mI2WlZlIvfC9wXETOBI8DK+nTLzMxqVUgCiog9EbG3TPxoRJxIq+OByG17O71sAcbmtwFfAr4CHBvo2JIEzAc2ptAGoGOQXTAzs2FqujEgSfMk7QZ2AjfnEhKSOoG3gHdICUTSHODDEfH9MtWNl9Ql6QVJHSl2IdCdq3cfUPGh65JWpTq6Dh48ONzumZlZ0rAEJOlpSbvKLIurlYuILRFxGXAVsEbS+Ny2dmAyMA6YL+l3gPuAf1OhuosiogR8EviqpEsAlTtslfasj4hSRJQmTZpUrelmZjYILQPvMjQRsWCY5fdIepdszKcrFz+WJhQsBn6Stj+XXVnjQ8BmSYsioisiDqQyr0t6DpgD/BXQJqklnQVNAw4Mp61mZjZ4TXUJLs1Oa0mvpwOzgDckTZA0OcVbgBuAVyLiVxExMSJmRMQM4AVgUUR0STpf0rhUZiJwDfByRATwLLAsHXYF8NgIdtPMzChuGvYSSfuAq4HH09gOwLXADknbgUeBWyLiEHAu2ZnNS8AOsnGgdQMc5lKgS9IOsoRzT0S8nLbdAdwu6TWyMaEH6tc7MzOrhbITAqtFqVSKrq6ugXc0M7P3SNqaxuNP01SX4MzM7OzhBGRmZoVwAjIzs0I4AZmZWSGcgMzMrBBOQGZmVggnIDMzK0TDbsVjZmbF2rRtP2s793Kgu4cpba2sbp9Fx5yK914ecU5AZmaj0KZt+1nzyE56jp8EYH93D2se2QnQNEnIl+DMzEahtZ1730s+vXqOn2RtZ79HsRXGCcjMbBQ60N0zqHgRnIDMzEahKW2tg4oXwQnIzGwUWt0+i9ZzxpwWaz1nDKvbZxXUov48CcHMbBTqnWjgWXBmZjbiOuZMbaqE05cvwZmZWSGcgMzMrBBOQGZmVggnIDMzK0QhCUjSckm7JZ2SVMrF50ranpYdkpbktj2ZYrslrZM0JsVvknQwV+4zuTIrJL2alhW5+MWStqT4w5LGjlTfzcwsU9QZ0C5gKfB8mXgpImYDC4H7JfXO1LsxIq4ELgcmActz5R6OiNlp+RaApAuAu4B5wFzgLknnp/3vBe6LiJnAEWBlvTtoZmbVFZKAImJPRPS7IVFEHI2IE2l1PBC5bW+nly3A2Py2CtqBpyLicEQcAZ4CFkoSMB/YmPbbAHQMtS9mZjY0TTcGJGmepN3ATuDmXEJCUifwFvAO7ycQgE9IeknSRkkfTrGpwJu5ffal2IVAd67e3nil9qyS1CWp6+DBg8PtnpmZJQ1LQJKelrSrzLK4WrmI2BIRlwFXAWskjc9tawcmA+PIzmIAvgfMiIiPAU+TndEAqFz1VeKV2rM+IkoRUZo0aVK1ppuZ2SA07E4IEbFgmOX3SHqXbMynKxc/JmkzsJjsEtsvc8W+STa+A9mZzXW5bdOA54BDQJuklnQWNA04MJy2mpnZ4DXVJbg0O60lvZ4OzALekDRB0uQUbwFuAF5J65NzVSwC9qTXncD1ks5Pkw+uBzojIoBngWVpvxXAY43tmZmZ9VXIveDS9Oqvkc1me1zS9nR57VrgTknHgVPALRFxSNIHgc2SxgFjgGeAdam62yQtAk4Ah4GbACLisKQvAS+m/b4YEYfT6zuAhyR9GdgGPNDYHpuZWV/KTgisFqVSKbq6ugbe0czM3iNpa0SU+sab6hKcmZmdPZyAzMysEE5AZmZWCCcgMzMrhBOQmZkVwgnIzMwK4QRkZmaFcAIyM7NCOAGZmVkhnIDMzKwQTkBmZlYIJyAzMyuEE5CZmRXCCcjMzArhBGRmZoVwAjIzs0I4AZmZWSGcgMzMrBCFJCBJyyXtlnRKUikXnytpe1p2SFqS2/Zkiu2WtE7SmBS/SdLBXLnP5MqczMU35+IXS9oi6VVJD0saO1J9NzOzTFFnQLuApcDzZeKliJgNLATul9SStt0YEVcClwOTgOW5cg9HxOy0fCsX78nFF+Xi9wL3RcRM4Aiwsm49MzOzmhSSgCJiT0TsLRM/GhEn0up4IHLb3k4vW4Cx+W2DIUnAfGBjCm0AOoZSl5mZDV3TjQFJmidpN7ATuDmXkJDUCbwFvMP7CQTgE5JekrRR0odz8fGSuiS9IKkjxS4EunP17gOmNqo/ZmZWXsMSkKSnJe0qsyyuVi4itkTEZcBVwBpJ43Pb2oHJwDiysxiA7wEzIuJjwNNkZzS9LoqIEvBJ4KuSLgFU7rBV+rEqJbGugwcPDtxxMzOrScMSUEQsiIjLyyyP1Vh+D/Au2ZhPPn4M2AwsTuu/jIjfpM3fBD6e2/dA+vo68BwwBzgEtOXGlqYBB6q0Y31ElCKiNGnSpFqabmZmNWiqS3BpdlpLej0dmAW8IWmCpMkp3gLcALyS1ifnqlgE7Enx8yWNS68nAtcAL0dEAM8Cy1KZFUBNSdHMzOqnZeBd6i9Nr/4a2Wy2xyVtT5fXrgXulHQcOAXcEhGHJH0Q2JwSyhjgGWBdqu42SYuAE8Bh4KYUv5RsFt0pskR7T0S8nLbdATwk6cvANuCBxvbYzMz6UnZCYLUolUrR1dVVdDPMzM4okram8fjTFHIGZGZmzW/Ttv2s7dzLge4eprS1srp9Fh1z6jdp2AnIzMz62bRtP2se2UnP8ZMA7O/uYc0jOwHqloSaahKCmZk1h7Wde99LPr16jp9kbWe/ewgMmROQmZn1c6C7Z1DxoXACMjOzfqa0tQ4qPhROQGZm1s/q9lm0njPmtFjrOWNY3T6rbsfwJAQzM+und6KBZ8GZmdmI65gzta4Jpy9fgjMzs0I4AZmZWSGcgMzMrBBOQGZmVggnIDMzK4Tvhj0Ikg4CPx9ksQ8Av6pTE+pR13DqGErZiWQPAbT6q+fPVtGarS8j3Z5GHq/edQ+lvukR0e+Jnk5ADSZpfUSsapa6hlPHUMpK6ip3G3Ybvnr+bBWt2foy0u1p5PHqXXc96/MluMb7XpPVNZw66tkXG77R9P1otr6MdHsaebx61123+nwGZA3lMyAzq8RnQNZo64tugJk1J58BmZlZIXwGZGZmhXACMjOzQjgBmZlZIZyAbERJ6pD0TUmPSbq+6PaYWXGcgGzYJD0o6S1Ju/rEF0raK+k1SXcCRMSmiPhj4CbgDwtorpk1CScgq4dvAwvzAUljgK8D/xj4KPDPJH00t8vn03YzO0s5AdmwRcTzwOE+4bnAaxHxekT8FngIWKzMvcATEfHTkW6rmTUPP5LbGmUq8GZufR8wD/hXwALgA5L+bkSsK6JxZlY8JyBrFJWJRUT8GfBnI90YM2s+vgRnjbIP+HBufRpwoKC2mFkTcgKyRnkRmCnpYkljgX8KbC64TWbWRJyAbNgk/TfgfwKzJO2TtDIiTgCfBTqBPcBfRsTuIttpZs3FNyM1M7NC+AzIzMwK4QRkZmaFcAIyM7NCOAGZmVkhnIDMzKwQTkBmZlYIJyAzMyuEE5CZmRXCCcisQJI+IukBSRuLbovZSHMCsooktUm6ZQSO8zeVjte7baQMtc+SWiX99/Qgvpql5yWt7FPXWEnPSyp7t3pJt0naI+k7g23ncNX6/gzjfSz7dN161D3UeoZzvHJPBbb3OQFZNW1Av1+89FC5uv3sRMQ/qHS83La6qKHt/dpQoz8CHomIkxWOe4Wk7/dZ/k65fdMD/H5I5UeW3wLcEBH/vM8x6vp9qaCN2t6fWvfr69v0ebpuHeseaj1DOl4NTwW2iPDipexC9hTTHmA78F2ym4p+A9gGTE/7bAK2AruBVSk2I+37zRT/AdAKnAs8DuwAdgF/mPb/dZnjrc1vS69vT+V2AX9S7Vh9+tG7z3ttL9fucm0APgX8JK3fD4yp8F79DTAjt74i1f8S8KMa3uuNfdavBP66zH7rgN8CO4F/XaFvld6nV4Bvpfh3yB4M+GPgVWBumWP1+35V+B6V+xkot1+t7+UMYFeNP5e9dffrc9F9Aa4GOnPra4A1Rf9eN9NSeAO8NO+S/0OQXp8CfrfPPhekr63pF/vCtO8JYHba9pfpF/YTwDdzZT+Qvv667/Fy+/Ru+zjZH91zgQnpD8ScSscq04/T2l6u3WX6fCnwPeCctP4N4F+UeZ/GAv83t34e8DIwNq23VXmPLyRLKj/L/3ECxgAHK5R5A5hYrm81vE9XkF352Ao8SPbgwMXApjLH6ff9qvA9qvQzsCu3T03vZaWfg2rbK/W56L4Ay4Bv5dY/Dfynon+vm2nxJTgbjJ9HxAt9YrdJ2gG8QPYAupkp/r8jYnt6vZXsl3gnsEDSvZJ+LyJ+NYhjXws8GhHvRsSvgUeA36tyrIHaXqndef+I7I/bi5K2p/WPlNlvItCdWz9J9gfsP0gqRUR3mTIARMQvI+LmiLgkIu7OxU8Cv5V0XqWyFfo20Pu0MyJOkf2R/mFkfxl3Uv49q/X7Vc/3ciiq9blXEX0p+1TgAXtzFvEjuW0w3s2vSLqO7DLO1RFxVNJzwPi0+Te5XU+SXRb7X5I+DtwA3C3pBxHxxRqPXe6XuVe/Y1Vr+wDt7nvMDRGxZoC29eTLpzovB/4JsF7StyLiGwPUUc444FgN++W/L7W+T6dy66co87eg3PcL+C/5fRrwXg5FtT4DhfXFTwUegM+ArJp3yC4nVfIB4Ej6Zf17wO9Wq0zSFOBoRPxX4N8Df38Qx3se6JD0tySdCywBflRDHwbb7nwbfggs650sIOkCSdP7VhYRR4Axksan/Wam/8YfAr5P+T9iVUm6kOwS3PFBFq3b+1Th+9X3e1Tpvey7X03vZY361j1gnwvqi58KPACfAVlFEfFLST9OU2L3lNnlSeBmSS8Be8kuW1RzBbBW0ingOPAvqxzviYhYndv2U0nfJhv4heza+jZJM4bQtYrt7tsG4PPAD9LssuPArcDPy9T5A7JLQU8Dn5N0NdmZyW7gj4fQxn8I/PVgC9X5fer3/arw/vR7L8t9LyUN+F4qe7rudcBESfuAuyLigT59LFd3vz4X3ZeIOCGp96nAY4AHw08FPo2fiGpWB5LmALdHxKfrVN8jZJMS9tajPrNm5EtwZnWQ/uN+VoP8IGo56XLNJicfG+18BmRmZoXwGZCZmRXCCcjMzArhBGRmZoVwAjIzs0I4AZmZWSGcgMzMrBBOQGZmVggnIDMzK8T/B5I3B/w5iAeoAAAAAElFTkSuQmCC\n", "text/plain": ["
"]}, "metadata": {"needs_background": "light"}, "output_type": "display_data"}], "source": ["bhm.ll_trans_scatter(bdata.models[2], 1,0)"]}, {"cell_type": "code", "execution_count": 43, "id": "bedeeb55", "metadata": {}, "outputs": [{"data": {"text/plain": [""]}, "execution_count": 43, "metadata": {}, "output_type": "execute_result"}, {"data": {"image/png": "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\n", "text/plain": ["
"]}, "metadata": {"needs_background": "light"}, "output_type": "display_data"}], "source": ["bhm.ll_trans_scatter(bdata.models[2], 1,0, rng=1.5, rng_only=False)"]}, {"cell_type": "raw", "id": "15d7ef00", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Forcing Recalculation\n", "=====================\n", "\n", "\n", "\n", "The :meth:`ModelError.Loglik_Error.get_E_err() ` / :meth:`ModelError.Loglik_Error.get_S_err() ` / :meth:`ModelError.Loglik_Error.get_trans_err() ` methods check if a value has been calculated, and can return such values without repeating the calculation.\n", "There are however lower level methods that calculate values regardless of whether they were calculated before, and over-write the value if it was.\n", "These are the :meth:`ModelError.Loglik_Error.E_eval() ` , :meth:`ModelError.Loglik_Error.S_eval() ` , and :meth:`ModelError.Loglik_Error.trans_eval() ` methods.\n", "\n", ".. note::\n", "\n", " Unlike the :attr:`ModelError.Loglik_Error.get_E/S/trans_err() ` methods, these methods do not have a return value, they just write their value to the corresponding array in the :class:`ModelError.Loglik_Error ` object.\n", "\n", "If these methods are called with no keyword arguments, then all values of E, S, or trans will be calculated, but if only specific values are to be calculated, you can specify the ``locs`` keyword argument.\n", "The ``locs`` keyword argument **must** be an iterable object where the elements are the desired states or transitions.\n", "\n", "These methods also accept the same ``thresh`` and ``flex`` keyword arguments, and come with the same warnings of setting them individually instead of globally as for :meth:`ModelError.Loglik_Error.get_E_err() ` / :meth:`ModelError.Loglik_Error.get_S_err() ` / :meth:`ModelError.Loglik_Error.get_trans_err() ` .\n", "Also present is the ``max_iter`` keyword argument which specifies how many iterations of the search algorithm to execute seeking a loglikelihood close to ``thresh`` less than the optimal, within a range of ``flex`` .\n", "Finally, for :meth:`ModelError.Loglik_Error.E_eval() ` / :meth:`ModelError.Loglik_Error.S_eval() ` , there is the ``step`` keyword argument, and for :meth:`ModelError.Loglik_Error.trans_eval() ` there is the ``factor`` argument, which specify the offset from the optimal value to use as an initial value in the search algorithm."]}, {"cell_type": "code", "execution_count": 44, "id": "06b5320c", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(data=[0.005624999999999991, 0.0019238281250000128, --],\n", " mask=[False, False, True],\n", " fill_value=nan)"]}, "execution_count": 44, "metadata": {}, "output_type": "execute_result"}], "source": ["# clear previously stored values so re-evaluation can be tested\n", "bdata.models[2].loglik_err.clear_all()\n", "\n", "bdata.models[2].loglik_err.E_eval(locs=(0,1))\n", "bdata.models[2].loglik_err.E"]}, {"cell_type": "code", "execution_count": 45, "id": "b2439b2f", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(data=[0.003408203125000009, 0.001953125, --],\n", " mask=[False, False, True],\n", " fill_value=-inf)"]}, "execution_count": 45, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.S_eval(locs=(0,1))\n", "bdata.models[2].loglik_err.S"]}, {"cell_type": "code", "execution_count": 46, "id": "3463da8c", "metadata": {}, "outputs": [{"data": {"text/plain": ["masked_array(\n", " data=[[--, 501.8588059030635, --],\n", " [193.29419084215726, --, --],\n", " [--, --, --]],\n", " mask=[[ True, False, True],\n", " [False, True, True],\n", " [ True, True, True]],\n", " fill_value=inf)"]}, "execution_count": 46, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.trans_eval(locs=((0,1), (1,0),))\n", "bdata.models[2].loglik_err.trans[:,:,0]"]}, {"cell_type": "raw", "id": "f391ef30", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Covariant optimizations\n", "***********************\n", "\n", "\n", "\n", "In the previous characterization of uncertainty, only a single parameter value could be varied at a time.\n", "Ideally we could make an N-dimensional grid of all combinations of parameter varied around their optimal values, but the shear number of points on such an N-dimensional grid grows so quickly with more states that such a method is not implemented here.\n", "\n", "Instead, we can take one parameter at a time, fix it's value to various offsets from the optimal, and find the optimal values for all other parameters.\n", "Thus we can see how changing one parameter affects the others, and how closely correlated they are to one another.\n", "\n", "This is done through the covariance methods: :meth:`ModelError.Loglik_Error.covar_trans() ` , :meth:`ModelError.Loglik_Error.covar_E() ` and :meth:`ModelError.Loglik_Error.covar_S() ` .\n", "Similar to the :meth:`ModelError.Loglik_Error.E_space ` / :meth:`ModelError.Loglik_Error.S_space ` / :meth:`ModelError.Loglik_Error.trans_space ` functions for the basic loglikelihood error, these functions take the particular transition/state you would like to evaluate as arguments, and also take the same ``rng`` and ``steps`` keyword arguments.\n", "\n", ".. note::\n", "\n", " Since these are optimization based parameters, these functions also take ``max_iter`` and ``converged_min`` keyword arguments to control when optimization terminates.\n", "\n", "Once finished, the results are stored in the corresponding index of :attr:`ModelError.Loglik_Error.trans_covar ` , :attr:`ModelError.Loglik_Error.E_covar ` , and :attr:`ModelError.Loglik_Error.S_covar ` , each of which is a numpy masked array of :class:`ModelError.ModelSet ` objects, so the calculated values will be stored in the corresponding index.\n", "As with the :attr:`ModelError.Bootstrap_Error.models ` attribute that was a :class:`ModelError.ModelSet ` object, access to the individual optimized parameters is provided through the appropriately named attributes."]}, {"cell_type": "code", "execution_count": 47, "id": "bac878eb", "metadata": {}, "outputs": [{"data": {"text/plain": ["The model converged after 82 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 78 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 74 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 66 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 53 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 54 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 68 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 76 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 80 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 84 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["array([0.15892296, 0.15906223, 0.15920173, 0.15934149, 0.15948151,\n", " 0.15962176, 0.15976239, 0.15990334, 0.16004464, 0.16018629])"]}, "execution_count": 47, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.covar_E(0)\n", "bdata.models[2].loglik_err.E_covar[0].E[:,1]"]}, {"cell_type": "code", "execution_count": 48, "id": "566037b4", "metadata": {}, "outputs": [{"data": {"text/plain": ["The model converged after 52 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 50 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 47 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 43 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 34 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 34 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 44 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 50 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 57 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 59 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["array([0.43215553, 0.43183874, 0.43152246, 0.43120671, 0.43089146,\n", " 0.4305768 , 0.43026259, 0.4299489 , 0.42963572, 0.42932306])"]}, "execution_count": 48, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.covar_S(1)\n", "bdata.models[2].loglik_err.S_covar[1].S[:,0]"]}, {"cell_type": "code", "execution_count": 49, "id": "62e03af6", "metadata": {}, "outputs": [{"data": {"text/plain": ["The model converged after 76 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 77 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 80 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 83 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 80 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 92 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 136 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["The model converged after 211 iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["Optimization reached maximum number of iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["Optimization reached maximum number of iterations"]}, "metadata": {}, "output_type": "display_data"}, {"data": {"text/plain": ["array([ 158.70168905, 207.5797375 , 271.51158681, 355.13361109,\n", " 464.51012722, 607.57318246, 794.6977911 , 1039.45433638,\n", " 1359.59270244, 1778.329506 ])"]}, "execution_count": 49, "metadata": {}, "output_type": "execute_result"}], "source": ["bdata.models[2].loglik_err.covar_trans(0,1)\n", "bdata.models[2].loglik_err.trans_covar[0,1].trans[:,0,1]"]}, {"cell_type": "raw", "id": "c54bf8a4", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Plotting Covariance\n", "===================\n", "\n", "\n", "\n", "There are 3 functions for plotting the results of covariance: :func:`covar_trans_ll_scatter() ` , :func:`covar_E_ll_scatter() ` and :func:`covar_S_ll_scatter() ` .\n", "These plot the loglikelihood of the optimized models against the value of the fixed parameter, they take two arguments: the :class:`ModelError.Loglik_Error ` data, and the state to plot, or in the case of ``covar_ll_trans_scatter()`` , the particular transition, and thus 3 arguments are required, the :class:`ModelError.Loglik_Error ` data, ``from_state`` and ``to_state`` :\n", "\n", ".. note::\n", "\n", " This function is smart, and will take either the base :class:`H2MM_result ` or :class:`ModelError.Loglik_Error ` , if the former, it will automatically extract the :class:`ModelError.Loglik_Error ` object from the :attr:`H2MM_result.loglik_err ` attribute.\n", "\n", ".. note::\n", "\n", " If you have not already run :meth:`ModelError.Loglik_Error.covar_trans() ` / :meth:`ModelError.Loglik_Error.covar_E() ` / :meth:`ModelError.Loglik_Error.covar_S() ` for the given state/transition, these functions will automatically run these with default parameters.\n", "\n", ""]}, {"cell_type": "code", "execution_count": 50, "id": "9b76444f", "metadata": {}, "outputs": [{"data": {"text/plain": [""]}, "execution_count": 50, "metadata": {}, "output_type": "execute_result"}, {"data": {"image/png": "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\n", "text/plain": ["
"]}, "metadata": {"needs_background": "light"}, "output_type": "display_data"}], "source": ["bhm.covar_trans_ll_scatter(bdata.models[2].loglik_err, 0,1)"]}, {"cell_type": "raw", "id": "95e883b0", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["These functions all also take the standard ``ax`` keyword argument, and all additional keyword arguments are passed to `ax.scatter() `_ "]}, {"cell_type": "code", "execution_count": 51, "id": "c34d2c15", "metadata": {}, "outputs": [{"data": {"text/plain": [""]}, "execution_count": 51, "metadata": {}, "output_type": "execute_result"}, {"data": {"image/png": "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\n", "text/plain": ["
"]}, "metadata": {"needs_background": "light"}, "output_type": "display_data"}], "source": ["fig, ax = plt.subplots()\n", "bhm.covar_E_ll_scatter(bdata.models[2], 0, ax=ax, s=60, marker='+', c='r')"]}, {"cell_type": "raw", "id": "82d62956", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Now there are a lot more values to plot, for instance, it is nice to see how when one parameter is offset, how does that affect the optimal values of other parameters.\n", "There are no built-in plotting functions for this, since specifying which parameter makes the function signatures rather long and awkward, so such plots are left to the user.\n", "Thankfully, if you know which particular combination of parameters you want to plot, it is not very difficult to setup your custom plots. \n", "\n", "This will also be a good way to teach how to access these different values.\n", "\n", "So say we want to see how, when the FRET efficiency of state 0 is varied, how does that affect the transitions from state 0 to state 1?\n", "\n", "To do this we need to extract these parameters, and then plot them.\n", "\n", "For demonstration purposes, we'll do this in 3 steps\n", "\n", "1. Isolate the covariance results ( :class:`ModelError.ModelSet ` object) of state 0.\n", "2. Extract the (fixed) E values of state 0, and the (optimized) transition values from state 0 to state 1\n", "3. Plot covariance"]}, {"cell_type": "code", "execution_count": 52, "id": "403503f3", "metadata": {}, "outputs": [{"data": {"text/plain": ["Text(0, 0.5, '0->1 $s^{-1}$, (optimized)')"]}, "execution_count": 52, "metadata": {}, "output_type": "execute_result"}, {"data": {"image/png": "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\n", "text/plain": ["
"]}, "metadata": {"needs_background": "light"}, "output_type": "display_data"}], "source": ["# get the ModelSet of state 0 E_covar\n", "covar_E_state0 = bdata.models[2].loglik_err.E_covar[0]\n", "\n", "# get fixed E values for x and optimized \n", "# note: we use ':' for the 0th dimension, so we look at each model\n", "# then specify the state/transition of interest\n", "x = covar_E_state0.E[:,0]\n", "y = covar_E_state0.trans[:,0,1]\n", "\n", "# plot\n", "plt.scatter(x,y)\n", "\n", "plt.xlabel(\"E$_{raw}$ state 0 (fixed)\")\n", "plt.ylabel(r\"0->1 $s^{-1}$, (optimized)\")"]}, {"cell_type": "raw", "id": "1c70a775", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Of course, we could just do this all in one line (plus axis labels): "]}, {"cell_type": "code", "execution_count": 53, "id": "7676152f", "metadata": {}, "outputs": [{"data": {"text/plain": ["Text(0, 0.5, '0->1 $s^{-1}$, (optimized)')"]}, "execution_count": 53, "metadata": {}, "output_type": "execute_result"}, {"data": {"image/png": "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\n", "text/plain": ["
"]}, "metadata": {"needs_background": "light"}, "output_type": "display_data"}], "source": ["plt.scatter(bdata.models[2].loglik_err.E_covar[0].E[:,0], bdata.models[2].loglik_err.E_covar[0].trans[:,0,1])\n", "plt.xlabel(\"E$_{raw}$ state 0 (fixed)\")\n", "plt.ylabel(r\"0->1 $s^{-1}$, (optimized)\")"]}, {"cell_type": "raw", "id": "140c4ff3", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Of course, there is no need to restrict ourselves to having the x axis be the fixed parameter, we can see how they all vary with one another, and maybe we'll give the fixed parameter as a color argument: "]}, {"cell_type": "code", "execution_count": 54, "id": "235a9917", "metadata": {}, "outputs": [{"data": {"text/plain": ["Text(0, 0.5, 'E$_{raw}$ state 2 (optimized)')"]}, "execution_count": 54, "metadata": {}, "output_type": "execute_result"}, {"data": {"image/png": "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\n", "text/plain": ["
"]}, "metadata": {"needs_background": "light"}, "output_type": "display_data"}], "source": ["x = bdata.models[2].loglik_err.E_covar[0].E[:,1] # state 1 E values\n", "y = bdata.models[2].loglik_err.E_covar[0].E[:,2] # state 2 E values\n", "c = bdata.models[2].loglik_err.E_covar[0].E[:,0] # state 0 E values (fixed)\n", "c /= c.max() # normalize values since using cmap\n", "plt.scatter(x, y, c=c, cmap='viridis')\n", "plt.xlabel(\"E$_{raw}$ state 1 (optimized)\")\n", "plt.ylabel(\"E$_{raw}$ state 2 (optimized)\")"]}, {"cell_type": "raw", "id": "aab1cbe6", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["Finally, let's think about combining the results of different covariances, we can use a 3-D plot to show how loglikelihood changes with respect to two parameters. \n", "\n", "So we'll take the covariance of E in state 0, and the covariance of S in state 1, plotting those values with E state 0 and S state 1 on the x and y axis, that way we can see the shape of two fixed axes with 1 free axis together."]}, {"cell_type": "code", "execution_count": 55, "id": "b359bf33", "metadata": {}, "outputs": [{"data": {"text/plain": [""]}, "execution_count": 55, "metadata": {}, "output_type": "execute_result"}, {"data": {"image/png": "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\n", "text/plain": ["
"]}, "metadata": {"needs_background": "light"}, "output_type": "display_data"}], "source": ["xE = bdata.models[2].loglik_err.E_covar[0].E[:,0] # state 0 E values\n", "yE = bdata.models[2].loglik_err.E_covar[0].S[:,1] # state 1 E values\n", "zE = bdata.models[2].loglik_err.E_covar[0].loglik # loglikelihood, note that there is no state specification\n", "\n", "xS = bdata.models[2].loglik_err.S_covar[1].E[:,0] # state 0 E values\n", "yS = bdata.models[2].loglik_err.S_covar[1].S[:,1] # state 1 E values\n", "zS = bdata.models[2].loglik_err.S_covar[1].loglik # loglikelihood, note that there is no state specification\n", "\n", "fig = plt.figure(figsize=(10,10))\n", "ax = fig.add_subplot(projection='3d')\n", "\n", "ax.scatter(xE, yE, zE, label='fixed: E$_{raw}$ state 0')\n", "ax.scatter(xS, yS, zS, label='fixed: S$_{raw}$ state 1')\n", "\n", "ax.set_xlabel(\"E$_{raw}$ state 0\")\n", "ax.set_ylabel(\"S$_{raw}$ state 1\")\n", "ax.set_zlabel(\"LL\")\n", "ax.legend()"]}, {"cell_type": "raw", "id": "1d6a4fbf", "metadata": {"raw_mimetype": "text/restructuredtext"}, "source": ["The key point, is that you have access to all non-lifetime model parameters (not dwell parameters) through identically named elements, and you can see how they correlate with one another and the overall logliklihood. ", "\n", "\n", "Download this documentation as a jupyter notebook here: :download:`Uncertainty.ipynb `"]}], "metadata": {"kernelspec": {"display_name": "Python 3 (ipykernel)", "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": "3.9.12"}}, "nbformat": 4, "nbformat_minor": 5}