{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"toc": "true"
},
"source": [
"Table of Contents
\n",
""
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# quick start "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Check relation betweeen dynamics and generalization. Hypothesis to make: generalization can only be understood in level of joint dynamical system, there is a clear link between the two**\n",
"\n",
"**Dynamnics determines generalization , not decoding , same decoding level can have very different dynamics , thus different generalization level. Only when the dynamics of RNN forms object correspond to real relevant objects for game, the generalization can be good. For instance , in a varying size game, you extend the size of game from 10 to 30, what will happen? You can do a kind of dynamical programing , according to which wall you have seen and how many steps you have passed , you decide future action. This can be achieved robustly by the dynamical system where fix points are correspond to walls**"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Resources\n",
"\n",
"* [*The* Reinforcement learning book from Sutton & Barto](http://incompleteideas.net/sutton/book/the-book-2nd.html)\n",
"* [The REINFORCE paper from Ronald J. Williams (1992)](http://www-anw.cs.umass.edu/~barto/courses/cs687/williams92simple.pdf)\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# FULL MODEL"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Packages"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Populating the interactive namespace from numpy and matplotlib\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/tie/anaconda3/lib/python3.6/site-packages/IPython/core/magics/pylab.py:160: UserWarning: pylab import has clobbered these variables: ['random']\n",
"`%matplotlib` prevents importing * from pylab and numpy\n",
" \"\\n`%matplotlib` prevents importing * from pylab and numpy\"\n"
]
}
],
"source": [
"%load_ext autoreload\n",
"%autoreload 2\n",
"\n",
"import numpy as np\n",
"from itertools import count\n",
"import random\n",
"\n",
"import torch\n",
"import torch.nn as nn\n",
"import torch.nn.functional as F\n",
"import torch.optim as optim\n",
"import torch.autograd as autograd\n",
"from torch.autograd import Variable\n",
"from torch.nn import init\n",
"from torch.nn import DataParallel\n",
"from torch.utils.data import DataLoader\n",
"\n",
"import matplotlib.mlab as mlab\n",
"import matplotlib.pyplot as plt\n",
"import matplotlib.animation\n",
"import seaborn as sns\n",
"from IPython.display import HTML\n",
"\n",
"import pretrain\n",
"from pretrain import *\n",
"\n",
"import navigation2\n",
"from navigation2 import *\n",
"\n",
"import sklearn\n",
"from sklearn.svm import SVC\n",
"from sklearn.manifold import Isomap\n",
"from sklearn.manifold import SpectralEmbedding\n",
"from sklearn.manifold import TSNE\n",
"from sklearn.linear_model import Lasso\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.metrics import f1_score\n",
"from sklearn.metrics import r2_score\n",
"from sklearn.model_selection import cross_validate\n",
"\n",
"import scipy\n",
"from scipy.spatial import distance\n",
"from scipy import signal\n",
"\n",
"import dynamics\n",
"from dynamics import*\n",
"\n",
"%pylab inline\n",
"import warnings\n",
"warnings.filterwarnings('ignore')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"For the liquid state approach to work, you need a lot of neurons as surplus or enough hidden to hidden connectivity to make it have an effect."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## POMDP RNN Game\n",
"\n",
"In this game , we use a new reward function determined by game, if the agent achieves the goal before 50, reward is 1. If time pass 50 reward is 0.5, once time pass 100 agent gets a reward of -0.5 . Practically, this is found to be easier to learn than the rewards as a continous function of time. Tf the agent learns to search in a efficient way, the largest possible way for search is to firstly arrive at corner then goes to the goal, which, takes about 50 steps, it is reasonble to make 50 and 100 as milestone thing. Also in principe as the game doesn't have a timer , it is not if it can use a reward as funtion of time"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"3 condition for ending , when pass time limit, game over"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"For weight update, it seems to be better do it after episode, as it makes non-sense evaluate strategy during episode, but a the end. Also, it is much quicker. "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"A programming of MDP here, hidden state is as state of enviroment"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Training from zero seems to be better because it will allow the agent to explore from new"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**A coherent result here is threhold kind behaviour of decode vs performance, after decode smaller than 10, the performance rises. The resutls is showeing that the learning rate vs performance doesn't characterize the threshold kind behaviour of performance change. It is only after lr = -5, the performance begins to rise , while the learning rate - performance curve creates a fake trends for first few points. This is not shown by explaination factor calculated by covariance. **"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"If look carefully, the slow timescale here is a very long transient, it will not return to baseline in a very long run of game, that is different from gru, which maintains a baseline. If the system is a true integrator, it should not loose the baseline, because there is no concept of return in this case. \n",
"\n",
"The very slow timescale could make the system harder to reset, except I use a strong input. So reset upon input and remember is always contradictory. If I design a task also need to pay attention to fast timescale , this will fail more throughly "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Real Game Dynamics analysis \n",
"The real game dynamcis shows huge difference performance in speed, stability of results, and also statistics of strategy "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**What is the relation to head direction, this order kept in limit cycle**\n",
"**Could we define the order parameter for behaviour and link it to dynamics?**"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Game Performance analysis"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Performance analysis\n",
"differ by speed, stability and final results on different sizes of arena, both for intra, and extrapolation"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### analysis on early stage\n",
"**much faster learning speed of pretrained net on small size of arena, compare resutls on the first iteration for small and large size of arena. That rise in speed in small size is not necessarlily related to decoding, but should relie more dynamics **"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"class PretrainTest():\n",
" def __init__(self, weight_write, holes = 0, inputs_type = (1, 0)):\n",
" self.pregame = PretrainGame(grid_size = (15, 15), holes = holes, random_seed = 4 , set_reward = [(0.5, 0.25), (0.5, 0.75)], input_type = inputs_type[0])\n",
" self.game = ValueMaxGame(grid_size = (15, 15), holes = 0, random_seed = 4 , set_reward = [(0.5, 0.25), (0.5, 0.75)], input_type = inputs_type[1])\n",
" self.weight = weight_write\n",
" \n",
" def loadweight(self, weight_load):\n",
"# need to take the state dict as a new dict for updating \n",
" net_dict = torch.load(weight_load)\n",
" list_modules = [('h2h', net_dict['h2h']), ('a2h', net_dict['a2h']), ('i2h', net_dict['i2h']), ('r2h', net_dict['r2h']), ('bh', net_dict['bh'])]\n",
" select_dict = OrderedDict(list_modules)\n",
" net = self.pregame.net.state_dict()\n",
" net.update(select_dict)\n",
" self.pregame.net.load_state_dict(net)\n",
" torch.save(self.pregame.net.state_dict(), self.weight) \n",
" \n",
" \n",
" def pretrain(self, trial, weight = None, lr = 1e-5, pretrain = True): \n",
" # start a pretrained game \n",
" self.pregame.net.cuda()\n",
" if pretrain == True:\n",
" lr = float(lr)\n",
" if weight != None:\n",
" self.pregame.net.load_state_dict(torch.load(weight))\n",
" self.pregame.fulltrain(lr_rate = lr, trials = int(1e3), batchsize = 4)\n",
" print ('pretrain end', torch.norm(self.pregame.net.h2h))\n",
" if pretrain == True:\n",
" torch.save(self.pregame.net.state_dict(), self.weight[:-1]+'{}'.format(trial))\n",
" else:\n",
" torch.save(self.pregame.net.state_dict(), self.weight+'{}'.format(trial))\n",
" if pretrain == True and trial <= 10:\n",
" self.weight = self.weight[:-1]+'{}'.format(trial)\n",
" elif pretrain == True and trial > 10:\n",
" self.weight = self.weight[:-2]+'{}'.format(trial)\n",
" elif pretrain == False:\n",
" self.weight = self.weight +'{}'.format(trial)\n",
" \n",
" def decode(self, weight = None):\n",
" if weight != None:\n",
" self.pregame.net.load_state_dict(torch.load(weight))\n",
" else:\n",
" self.pregame.net.load_state_dict(torch.load(self.weight))\n",
" rls_q = RLS(1e2)\n",
" rls_sl = RLS(1e2)\n",
" self.game.net.cpu()\n",
" self.game.experiment(rls_q, rls_sl,20, epsilon = 0.5, train_hidden = False, train_q = False, size_range=(15, 16), test = True) \n",
" def precision():\n",
" prec0 = np.mean(decodetest(self.game, reward_control= 0, epsilon = 0)[0] + decodetest(self.game, reward_control = 1, epsilon = 0)[0])\n",
" prec1 = np.mean(decodetest(self.game, reward_control= 0, epsilon = 0.5)[0] + decodetest(self.game, reward_control = 1, epsilon = 0.5)[0])\n",
" prec2 = np.mean(decodetest(self.game, reward_control= 0, epsilon = 1)[0] + decodetest(self.game, reward_control = 1, epsilon = 1)[0])\n",
" return (prec0 + prec1 + prec2)/3, (prec2 - prec0)\n",
" print ('decode train finish')\n",
" Prec, dif = precision()\n",
" print ('decode end', Prec, 'dif decode', dif)\n",
" return Prec, dif\n",
" # q learning session \n",
" \n",
" \n",
" def qlearn(self, weight_read, weight_write, iterations = 5, save = True, size_train = np.arange(10, 51, 10), \\\n",
" size_test = [10, 30], train_only = False, test_only = False, noise = 0.3, h2o = True,\n",
" k_action = 1, k_internal = 1, k_stim = 1):\n",
" self.game.net.load_state_dict(torch.load(weight_read))\n",
" if h2o == True:\n",
" self.game.net.h2o = nn.Parameter(torch.randn(512, 4) * 0.01 * np.sqrt(2.0/(512 + 4)))\n",
" self.game.net.k_action = k_action\n",
" self.game.net.k_internal = k_internal\n",
" self.game.net.k_stim = k_stim\n",
" e_rate = [noise for r in range(iterations)] \n",
" rls_q = RLS(1e2)\n",
" rls_sl = RLS(1e2)\n",
" Rewards = []\n",
" # q leanring phase\n",
" for n,e in enumerate(e_rate):\n",
" prob = np.ones(len(size_train)) \n",
" prob = prob/np.sum(prob)\n",
" if test_only == False:\n",
" self.game.experiment(rls_q, rls_sl, iterations = 50, epochs= 10, epsilon = e, size_range = size_train) \n",
" if save == True:\n",
" torch.save(self.game.net.state_dict(), weight_write + '_{}'.format(n))\n",
" def testing(game):\n",
" Rewards00 = Test(game, reward_control = 0, size = size_test[0], test = 1)\n",
" Rewards01 = Test(game, reward_control = 1, size = size_test[0], test = 1)\n",
" rewards_s = (np.sum(Rewards00) + np.sum(Rewards01))/2\n",
" Rewards10 = Test(game, reward_control = 0, size = size_test[1], test = 2)\n",
" Rewards11 = Test(game, reward_control = 1, size = size_test[1], test = 2)\n",
" rewards_l = (np.sum(Rewards10) + np.sum(Rewards11))/2\n",
" return rewards_s, rewards_l\n",
" # load weight if test only is true \n",
" if test_only == True:\n",
" self.game.net.load_state_dict(torch.load(weight_write))\n",
" if train_only == False:\n",
" rewards_s, rewards_l = testing(self.game)\n",
" print (n, 'rewards_s', rewards_s, 'rewards_l', rewards_l)\n",
" Rewards.append((rewards_s, rewards_l))\n",
" return Rewards\n",
" \n",
" def TestAllSizes(self, size_range = np.arange(5, 50, 5), k_action = 1, k_internal = 1, \n",
" k_stim = 1, limit_set = 4, test_size = 1):\n",
" self.game.net.load_state_dict(torch.load(self.weight))\n",
" self.game.net.cpu()\n",
" self.Performance = []\n",
" self.game.net.k_action = k_action\n",
" self.game.net.k_internal = k_internal\n",
" self.game.net.k_stim = k_stim\n",
" for size in size_range:\n",
" Rewards0 = Test(self.game, reward_control = 0, size = size, limit_set = limit_set, test = test_size)\n",
" Rewards1 = Test(self.game, reward_control = 1, size = size, limit_set = limit_set, test = test_size)\n",
" self.Performance.append((Rewards0 + Rewards1)/2)"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [],
"source": [
"def performance_measure(size, test_size, trials = 300, \\\n",
" shuffle = False, pretrain = False, random = False, k_action = 1, k_internal = 1, k_stim = 1, limit_set = 4):\n",
" performance = np.zeros((trials, 5))\n",
" for iters1 in np.arange(trials):\n",
" for iters2 in np.arange(5, 10):\n",
" if shuffle == True:\n",
" weight = 'weights_2_shuffle/rnn_1515tanh512_checkpoint300_{}_{}'.format(iters1+3, iters2)\n",
" if random == True:\n",
" weight = 'weights_2_random/rnn_1515tanh512_checkpoint300_{}_{}'.format(iters1+3, iters2)\n",
" if pretrain == True:\n",
" weight = 'weights2/rnn_1515tanh512_checkpoint300_{}_{}'.format(iters1+3, iters2)\n",
" Pretest = PretrainTest(weight_write = weight)\n",
" Pretest.TestAllSizes(size_range = [size], limit_set = limit_set, test_size = test_size)\n",
" performance[iters1, iters2-5] = Pretest.Performance[0]\n",
" print(np.max(performance, axis = 1))\n",
" return performance\n",
"# performance_measure(0, 0, 10, 0)"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"[]"
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"## The rank of pertubation \n",
"Net = torch.load('weights_net1/rnn_1515tanh512_checkpoint39_0_9')\n",
"h2h_39 = Net['h2h'].clone()\n",
"Net = torch.load('weights_cpu1/rnn_1515tanh512_checkpoint0')\n",
"h2h_0 = Net['h2h'].clone()\n",
"dW = h2h_39 - h2h_0\n",
"dW = dW.data.numpy()\n",
"dW_shuffle = np.random.permutation(dW.ravel()).reshape(512, 512)\n",
"u, s, vh = np.linalg.svd(dW)\n",
"u_shuffle, s_shuffle, vh_shuffle = np.linalg.svd(dW_shuffle)\n",
"plt.figure()\n",
"plt.plot(s[:100],'r.')\n",
"plt.plot(s_shuffle[:100],'b.')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Learning History"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"trial = 300\n",
"\n",
"for iters1, noise in enumerate(3 * [0.0]):\n",
" performance = []\n",
" for iters2 in np.arange(11):\n",
" if iters2 == 0:\n",
" Pretest = PretrainTest(weight_write = 'weights_cpu/rnn_1515tanh512_checkpoint{}'.format(trial))\n",
" Pretest.TestAllSizes(size_range = [10], limit_set = 4, test_size = 0)\n",
" performance.append(Pretest.Performance)\n",
" if iters2>0:\n",
" Pretest = PretrainTest(weight_write = 'weights2/rnn_1515tanh512_checkpoint{}_{}_{}'.format(trial, iters1, iters2-1))\n",
" Pretest.TestAllSizes(size_range = [10], limit_set = 4, test_size = 0)\n",
" performance.append(Pretest.Performance)\n",
" np.save('Rewards_s_{}_{}.npy'.format(iters1, 39), performance)"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"([,\n",
" ,\n",
" ],\n",
" )"
]
},
"execution_count": 19,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAESCAYAAAAIfCk9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"ax = subplot(1, 1, 1)\n",
"ax.spines['top'].set_visible(False)\n",
"ax.spines['right'].set_visible(False)\n",
"Performance_pos = []\n",
"for iters, noise in enumerate(3 * [0.0]):\n",
" for trial in [39]: \n",
"# reward = np.load('Rewards_s_{}_{}_s.npy'.format(iters, trial))\n",
"# plt.plot(reward, 'c.--')\n",
" reward = np.load('Rewards_s_{}_{}.npy'.format(iters, trial))\n",
" plt.plot(reward, 'r.-')\n",
" Performance_pos.append(reward)\n",
"np.save('Performance_pos', Performance_pos)\n",
"# plt.xticks([0, 10, 20, 30], size = 15)\n",
"# plt.yticks([-0.4, 0, 0.4, 0.8], size = 15)\n",
"plt.title('scaling', size = 20)\n",
"plt.yticks([0, 0.25, 0.5, 0.75, 1], size = 15)\n",
"plt.xticks([0, 5, 10], size = 15)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Generalization"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[0.85458433 0.95430402 0.88056241]\n"
]
}
],
"source": [
"performance10_s = performance_measure(10, 0, trials = 3, k_action=1, k_internal=1, k_stim=1, shuffle = False, pretrain=True)"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [],
"source": [
"np.save('performance10', performance10_s)"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[0.96329207 0.96926579 0.97065962]\n"
]
}
],
"source": [
"performance10 = performance_measure(10, 0, trials = 3, k_action=1, k_internal=1, k_stim=1, pretrain = True)"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[ 0.38521771 0.52222303 -0.04578278]\n"
]
}
],
"source": [
"performance30 = performance_measure(30, 1, trials = 3, k_action=1, k_internal=1, k_stim=1, shuffle = False, pretrain=True)"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [],
"source": [
"np.save('performance30', performance)"
]
},
{
"cell_type": "code",
"execution_count": 34,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[0.35502959 0.43786982 0.10059172 0.10059172 0.31952663]\n"
]
}
],
"source": [
"performance50_s = performance_measure(50, 2, trials = 5, k_action=1, k_internal=1, k_stim=1, limit_set = 4, shuffle = True)"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[ 0.03957431 -0.03178359 -0.32431379]\n"
]
}
],
"source": [
"performance50 = performance_measure(50, 2, trials = 3, k_action=1, k_internal=1, k_stim=1, limit_set = 4, pretrain = True)"
]
},
{
"cell_type": "code",
"execution_count": 35,
"metadata": {},
"outputs": [],
"source": [
"np.save('performance50', performance50)"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[-0.04591837 -0.01530612 -0.24489796 -0.15306122 -0.04591837]\n"
]
}
],
"source": [
"performance70_s = performance_measure(70, 3, trials = 3, k_action=1, k_internal=1, k_stim=1, limit_set = 4, shuffle = True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[-0.04569866 -0.21776189 -0.41834027]\n"
]
}
],
"source": [
"performance70 = performance_measure(70, 3, trials = 3, k_action=1, k_internal=1, k_stim=1, limit_set = 4, pretrain = True)"
]
},
{
"cell_type": "code",
"execution_count": 36,
"metadata": {},
"outputs": [],
"source": [
"np.save('performance70', performance70)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"performance90 = performance_measure(90, 4, trials = 3, k_action=1, k_internal=1, k_stim=1, limit_set = 4, pretrain = True)"
]
},
{
"cell_type": "code",
"execution_count": 37,
"metadata": {},
"outputs": [],
"source": [
"np.save('performance90', performance90)"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[-0.15976331 -0.30177515 -0.46153846 -0.44378698 -0.20118343]\n"
]
}
],
"source": [
"performance90_s_f = performance_measure(90, 4, trials = 3, k_action=1, k_internal=1, k_stim=1, limit_set = 4, shuffle = True)"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[-0.07100592 -0.34911243 -0.0591716 -0.20118343 -0.33136095]\n"
]
}
],
"source": [
"performance90_f = performance_measure(90, 4, trials = 3, k_action=1, k_internal=1, k_stim=1, limit_set = 4, pretrain = True)"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [],
"source": [
"np.save('performance_10', performance10)\n",
"np.save('performance_30', performance30)\n",
"# np.save('performance10_shuffle', performance10_s)\n",
"# np.save('performance50_shuffle', performance50_s)\n",
"np.save('performance_50', performance50)\n",
"# np.save('performance70_shuffle', performance70_s)\n",
"np.save('performance_70', performance70)\n",
"# np.save('performance90_shuffle_f', performance90_s_f)\n",
"np.save('performance_90', performance90)"
]
},
{
"cell_type": "code",
"execution_count": 38,
"metadata": {},
"outputs": [
{
"ename": "FileNotFoundError",
"evalue": "[Errno 2] No such file or directory: 'performance30_shuffle.npy'",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mFileNotFoundError\u001b[0m Traceback (most recent call last)",
"\u001b[0;32m\u001b[0m in \u001b[0;36m\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mperformance30_s\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mload\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'performance30_shuffle.npy'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 2\u001b[0m \u001b[0mperformance_g10_s\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mperformance10_s\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 3\u001b[0m \u001b[0mperformance_g30_s\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mperformance30_s\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 4\u001b[0m \u001b[0mperformance_g50_s\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mperformance50_s\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 5\u001b[0m \u001b[0mperformance_g70_s\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mperformance70_s\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/numpy/lib/npyio.py\u001b[0m in \u001b[0;36mload\u001b[0;34m(file, mmap_mode, allow_pickle, fix_imports, encoding)\u001b[0m\n\u001b[1;32m 370\u001b[0m \u001b[0mown_fid\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 371\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfile\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbasestring\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 372\u001b[0;31m \u001b[0mfid\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfile\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"rb\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 373\u001b[0m \u001b[0mown_fid\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 374\u001b[0m \u001b[0;32melif\u001b[0m \u001b[0mis_pathlib_path\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfile\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;31mFileNotFoundError\u001b[0m: [Errno 2] No such file or directory: 'performance30_shuffle.npy'"
]
}
],
"source": [
"performance30_s = np.load('performance30_shuffle.npy') \n",
"performance_g10_s = np.max(performance10_s, axis = 1)\n",
"performance_g30_s = np.max(performance30_s, axis = 1)\n",
"performance_g50_s = np.max(performance50_s, axis = 1) \n",
"performance_g70_s = np.max(performance70_s, axis = 1) \n",
"performance_g90_s = np.max(performance90_s_f, axis = 1)"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [],
"source": [
"performance_10 = np.load('performance_10.npy')\n",
"performance_30 = np.load('performance_30.npy')\n",
"performance_50 = np.load('performance_50.npy') \n",
"performance_70 = np.load('performance_70.npy') \n",
"performance_90 = np.load('performance_90.npy')\n",
"performance_g10 = np.max(performance_10, axis = 1)\n",
"performance_g30 = np.max(performance_30, axis = 1)\n",
"performance_g50 = np.max(performance_50, axis = 1) \n",
"performance_g70 = np.max(performance_70, axis = 1) \n",
"performance_g90 = np.max(performance_90, axis = 1)"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [],
"source": [
"# PG_s = [performance_g10_s, performance_g30_s, performance_g50_s, performance_g70_s, performance_g90_s]\n",
"# pG_s = np.concatenate(PG_s)\n",
"PG = [performance_g10, performance_g30, performance_g50, performance_g70, performance_g90]\n",
"pG = np.concatenate(PG)"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [],
"source": [
"np.save('PG_Scaling_PosNet_2', PG)"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
""
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEBCAYAAABrF5JMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZOT/AZx9BhARxK15AAAAAElFTkSuQmCC\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"ax = subplot(1, 1, 1)\n",
"ax.spines['top'].set_visible(False)\n",
"ax.spines['right'].set_visible(False)\n",
"plt.xticks(size = 15)\n",
"plt.yticks([-0.6, -0.2, 0.2, 0.6, 1], size = 15)\n",
"plt.errorbar(np.arange(10, 91, 20), np.mean(PG, axis = 1), yerr = np.std(PG, axis = 1), color = 'r')\n",
"# plt.errorbar(np.arange(10, 91, 20), np.mean(PG_s, axis = 1), yerr = np.std(PG_s, axis = 1), color = 'c')\n"
]
},
{
"cell_type": "code",
"execution_count": 58,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[ 0.63905325 0.27810651 -0.00591716 -0.08284024 0.20710059]\n"
]
}
],
"source": [
"performance = performance_measure(90, 4, trials = 5, k_action=1, k_internal=1, k_stim=1, limit_set = 16)"
]
},
{
"cell_type": "code",
"execution_count": 68,
"metadata": {},
"outputs": [],
"source": [
"np.save('performance90_shuffle', performance)"
]
},
{
"cell_type": "code",
"execution_count": 81,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[ 0.07100592 -0.28994083 0.3964497 0.18343195 -0.25443787]\n"
]
}
],
"source": [
"performance = performance_measure(90, 4, trials = 5, k_action=1, k_internal=1, k_stim=1, limit_set = 16,\n",
" pretrain = True)"
]
},
{
"cell_type": "code",
"execution_count": 82,
"metadata": {},
"outputs": [],
"source": [
"np.save('performance90', performance)"
]
},
{
"cell_type": "code",
"execution_count": 83,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[0.76 0.28 0.58 0.74 0.52]\n"
]
}
],
"source": [
"performance = performance_measure(30, 1, trials = 5, k_action=1, k_internal=1, k_stim=1, pretrain = True)"
]
},
{
"cell_type": "code",
"execution_count": 84,
"metadata": {},
"outputs": [],
"source": [
"np.save('performance30', performance)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Performance 30 as reference"
]
},
{
"cell_type": "code",
"execution_count": 77,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"0.6187999999999999"
]
},
"execution_count": 77,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"perform30 = np.load('performance30_shuffle.npy')\n",
"np.mean(perform30[perform30>0.0]) "
]
},
{
"cell_type": "code",
"execution_count": 85,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"0.45875"
]
},
"execution_count": 85,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"perform30 = np.load('performance30.npy')\n",
"np.mean(perform30[perform30>0.0]) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Performance 90 as test"
]
},
{
"cell_type": "code",
"execution_count": 78,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"-0.08001029071263185"
]
},
"execution_count": 78,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"perform90 = np.load('performance90_shuffle.npy')\n",
"np.mean(perform90[perform30>0.3])"
]
},
{
"cell_type": "code",
"execution_count": 86,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"-0.1643655489809336"
]
},
"execution_count": 86,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"perform90 = np.load('performance90.npy')\n",
"np.mean(perform90[perform30>0.3])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": 45,
"metadata": {},
"outputs": [
{
"ename": "TypeError",
"evalue": "'numpy.int64' object is not iterable",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mTypeError\u001b[0m Traceback (most recent call last)",
"\u001b[0;32m\u001b[0m in \u001b[0;36m\u001b[0;34m()\u001b[0m\n\u001b[1;32m 2\u001b[0m \u001b[0mpca\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpca\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mT_duration\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m5\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 3\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m \u001b[0mpca\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mDynamics\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mActions\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m3\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlegend\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mT_total\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m200\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mT_stim\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m100\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mT_duration\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m3\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreadout_random\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mopen_loop\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 5\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 6\u001b[0m \u001b[0;31m# long dynamics for different stimulus\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m~/Research/PhD/NavigationPaper_606/pretrain/pretrain_pos_entropy_ml_v0/dynamics.py\u001b[0m in \u001b[0;36mDynamics\u001b[0;34m(self, T_total, T_stim, T_duration, Hiddens, noise, iters, Actions, e, same, legend, corner, open_loop, readout_random, h2o)\u001b[0m\n\u001b[1;32m 234\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mStim\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstack\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mStim1\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0mStim2\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0mStim3\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 235\u001b[0m \u001b[0;31m# trace in empty room\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 236\u001b[0;31m \u001b[0mPos1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhidden1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdh1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mactions\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtrajectory_empty\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpos0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgame\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mStim\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maction_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0maction\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0me\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mopen_loop\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mopen_loop\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 237\u001b[0m \u001b[0mT_transient\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdh1\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mT_stim\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m>\u001b[0m\u001b[0;36m1e-1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 238\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mTs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mT_transient\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m~/Research/PhD/NavigationPaper_606/pretrain/pretrain_pos_entropy_ml_v0/dynamics.py\u001b[0m in \u001b[0;36mtrajectory_empty\u001b[0;34m(pos0, game, Stim, reward_control, action_, e, open_loop, init_hidden, hidden)\u001b[0m\n\u001b[1;32m 116\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mstim\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mStim\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 117\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mopen_loop\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 118\u001b[0;31m \u001b[0mdone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maction\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgame\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep_empty\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstim\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mstim\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maction_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0maction_\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mepsilon\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mopen_loop\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mopen_loop\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# Down\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 119\u001b[0m \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 120\u001b[0m \u001b[0mdone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maction\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgame\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep_empty\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstim\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mstim\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maction_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0maction\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mepsilon\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mopen_loop\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mopen_loop\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# Down\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;31mTypeError\u001b[0m: 'numpy.int64' object is not iterable"
]
}
],
"source": [
"pca = PCA(weight = 'weights2/rnn_1515tanh512_checkpoint399_0_9')\n",
"pca.pca(T_duration = 5)\n",
"\n",
"pca.Dynamics(Actions = [0, 1, 2, 3], legend = True, T_total = 200, T_stim = 100, T_duration = 3, readout_random = False, open_loop = True)\n",
"\n",
"# long dynamics for different stimulus\n",
"colors = ['b', 'g', 'r', 'm', 'c']\n",
"for i in range(4):\n",
" c = colors[i]\n",
" for j in range(1):\n",
" plt.plot(pca.PCs[i, j], color = c)\n",
"plt.xticks([])\n",
"plt.yticks([])\n",
"plt.figure()\n",
"for i in range(4):\n",
" for j in range(4):\n",
" c = colors[j]\n",
" plt.plot(pca.PCs[i, j], color = c)"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAYYAAAD8CAYAAABzTgP2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nlFJKBdHEoJRSKogmBqWUUkH+H1KPZ01xgdMOAAAAAElFTkSuQmCC\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAA60AAAAtCAYAAABF/Y2aAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAD8CAYAAACfF6SlAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"pca = PCA(weight = 'weights_net1/weights2/rnn_1515tanh512_checkpoint39_0_0')\n",
"pca.pca(T_duration = 5)\n",
"\n",
"pca.Dynamics(Actions = [0, 1, 2, 3], legend = True, T_total = 200, T_stim = 100, T_duration = 3, readout_random = False, open_loop = True)\n",
"\n",
"# long dynamics for different stimulus\n",
"colors = ['b', 'g', 'r', 'm', 'c']\n",
"for i in range(4):\n",
" c = colors[i]\n",
" for j in range(1):\n",
" plt.plot(pca.PCs[i, j], color = c)\n",
"plt.figure()\n",
"plt.matshow(pca.Actions[0][0].reshape(1, -1))\n",
"plt.figure()\n",
"for i in range(4):\n",
" for j in range(4):\n",
" c = colors[j]\n",
" plt.plot(pca.PCs[i, j], color = c)"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-0.31418962623535845 -0.4477606501739859\n",
"0.419493464147394 -0.674066490294598\n",
"0.19372434640345315 -0.5096941120851743\n",
"-0.019540488982937504 -0.2893080186203524\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAD8CAYAAACfF6SlAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nib9CoVDUIP8/ddoqezgkXHkAAAAASUVORK5CYII=\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAA60AAAAtCAYAAABF/Y2aAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSZIkqbUsWiVJkiRJrfU/Zxg2UOtZcOIAAAAASUVORK5CYII=\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD8CAYAAAB0IB+mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n4SUk4AshhJf4//2dtptUrkOoAAAAAElFTkSuQmCC\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAD8CAYAAACfF6SlAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nf4ZhGB3C4s8wDKND/n/Yjbz8LBwWoQAAAABJRU5ErkJggg==\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAA60AAAAtCAYAAABF/Y2aAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nVdUrZ8/IJDgAAAAASUVORK5CYII=\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD8CAYAAAB0IB+mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD8CAYAAAB+UHOxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nqxV8ZcPAnHYAAAAASUVORK5CYII=\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAA60AAAAtCAYAAABF/Y2aAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJEmSJPUti1ZJkiRJUt/6H2vlHXVlbcI3AAAAAElFTkSuQmCC\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD8CAYAAAB0IB+mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAD8CAYAAACfF6SlAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxin4VAlNPuIAAAAASUVORK5CYII=\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAA60AAAAtCAYAAABF/Y2aAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD8CAYAAAB+UHOxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYpQGAKWUilH/HwPwthyjq+qoAAAAAElFTkSuQmCC\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# reference from net 1 , It is surprising how network reconciles stimulus memroy and path integration\n",
"pca = PCA(weight = 'weights_net1/weights2/rnn_1515tanh512_checkpoint39_0_9')\n",
"pca.pca(T_duration = 5)\n",
"pca.Dynamics(Actions = [0], legend = True, T_total = 200, T_stim = 100, T_duration = 3, \\\n",
" readout_random = False, open_loop = False, e = 1)\n",
"# long dynamics for different stimulus\n",
"colors = ['b', 'g', 'r', 'm', 'c'] \n",
"for i in range(4):\n",
" plt.figure()\n",
" plt.plot(pca.PCs[i, 0][100:]) \n",
" plt.figure()\n",
" plt.matshow(pca.Actions[i][0].reshape(1, -1))\n",
" plt.figure()\n",
" plt.plot(pca.Ys[i][0])\n",
" plt.plot(pca.Xs[i][0])\n",
" print (np.corrcoef(pca.Ys[i][0], pca.PCs[i, 0][100:])[0, 1], np.corrcoef(pca.Xs[i][0], pca.PCs[i, 0][100:])[0, 1])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## GLM"
]
},
{
"cell_type": "code",
"execution_count": 61,
"metadata": {},
"outputs": [],
"source": [
"def variance_decompose(pca, T, open_loop = True, alpha = 1e-4):\n",
" Actions = np.zeros((pca.Hiddens.shape[0], pca.Hiddens.shape[1], T))\n",
" Y = np.ones((pca.Hiddens.shape[0], pca.Hiddens.shape[1], T))\n",
" X = np.ones((pca.Hiddens.shape[0], pca.Hiddens.shape[1], T))\n",
" Stims = np.ones((pca.Hiddens.shape[0], pca.Hiddens.shape[1], T))\n",
" colors = ['b', 'g', 'r', 'm']\n",
" for i in range(pca.Hiddens.shape[0]):\n",
" for j in range(pca.Hiddens.shape[1]):\n",
" if open_loop == False:\n",
" Actions[i, j] = pca.Actions[i, j, :T]\n",
" else:\n",
" Actions[i, j] = j\n",
" Stims[i, j] = i\n",
" Y[i, j] = pca.Ys[i, j, :T]\n",
" X[i, j] = pca.Xs[i, j, :T]\n",
"\n",
" x = pca.Hiddens[:, :, :T, :].reshape(-1, 512)\n",
" z = (x - np.min(x))/(np.max(x) - np.min(x))\n",
" y = np.log(z/(1-z + 1e-3) + 1e-3)\n",
" ## features:\n",
" A = np.array([np.eye(4)[int(a)] for a in Actions.reshape(-1)]).reshape(-1, 4)\n",
" S = np.array([np.eye(4)[int(s)] for s in Stims.reshape(-1)]).reshape(-1, 4)\n",
" Y = Y.reshape(-1, 1)/np.max(np.abs(Y))\n",
" X = X.reshape(-1, 1)/np.max(np.abs(X))\n",
" Features = np.concatenate((A, S, Y, X), axis = 1)\n",
" Features_A = np.concatenate((S, Y, X), axis = 1)\n",
" Features_S = np.concatenate((A, Y, X), axis = 1)\n",
" Features_Y = np.concatenate((A, S, X), axis = 1)\n",
" Features_X = np.concatenate((A, S, Y), axis = 1)\n",
"\n",
" clf = Lasso(alpha = alpha)\n",
" cv_scores_all = cross_validate(clf, Features, y, cv = 5, scoring=('r2'), return_train_score = True)\n",
" importances = []\n",
" importances_std = []\n",
" for feature in [Features_A, Features_S, Features_Y, Features_X]:\n",
" clf = Lasso(alpha= alpha)\n",
" cv_scores = cross_validate(clf, feature, y, cv = 5, scoring=('r2'), return_train_score = True)\n",
" print ('total variance', np.mean(cv_scores_all['train_score']), np.mean(cv_scores['train_score']), \\\n",
" np.mean(cv_scores_all['test_score']), np.mean(cv_scores['test_score']))\n",
" importances.append(np.mean(cv_scores_all['train_score'] - cv_scores['train_score']))\n",
" importances_std.append(np.std(cv_scores_all['train_score'] - cv_scores['train_score']))\n",
" return importances, importances_std\n",
"\n",
"def Memory(weight, shuffle = False):\n",
" Info_P = np.zeros(2)\n",
" Info_I = np.zeros(2)\n",
" Info_A = np.zeros(2)\n",
" Info_P_std = np.zeros(2)\n",
" Info_I_std = np.zeros(2)\n",
" Info_A_std = np.zeros(2)\n",
" pca = PCA(weight = weight)\n",
" pca.pca(T_duration = 3, shuffle = shuffle)\n",
" pca.Dynamics(Actions = 50 * [0], legend = True, T_total = 200, T_stim = 100, T_duration = 3, \\\n",
" readout_random = True, open_loop = False, e = 1)\n",
" for i in range(2):\n",
" T = 10 + i * 40\n",
" importances, importances_std = variance_decompose(pca, T, open_loop=False)\n",
" Info_A[i] = importances[0]\n",
" Info_I[i] = importances[1]\n",
" Info_P[i] = importances[2] + importances[3] \n",
" Info_A_std[i] = importances_std[0]\n",
" Info_I_std[i] = importances_std[1]\n",
" Info_P_std[i] = importances_std[2] + importances_std[3] \n",
" print (np.mean(Info_A), np.mean(Info_I), np.mean(Info_P))\n",
" return Info_A, Info_I, Info_P, Info_A_std, Info_I_std, Info_P_std \n",
"\n",
"# weight = 'weights_net1_shuffle/rnn_1515tanh512_checkpoint39_0_9'\n",
"# Info_A, Info_I, Info_P, Info_A_std, Info_I_std, Info_P_std = Memory(weight, k_action = 1, k_stim = 1, k_internal = 1.)\n",
"# print ('decay', Info_I[-1]/Info_I[0])"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"total variance 0.45869659896770687 0.4560207363153543 0.2693681850034667 0.2696644126383137\n",
"total variance 0.45869659896770687 0.17031593369356943 0.2693681850034667 -0.4497041854005424\n",
"total variance 0.45869659896770687 0.3531040750482831 0.2693681850034667 0.15281895717842417\n",
"total variance 0.45869659896770687 0.41432977343598587 0.2693681850034667 0.2233487344390989\n",
"total variance 0.3926204053521799 0.3915117328513492 0.21595534394735755 0.2156827207377176\n",
"total variance 0.3926204053521799 0.13406594189496024 0.21595534394735755 -0.3997735690013033\n",
"total variance 0.3926204053521799 0.3383501649649733 0.21595534394735755 0.14751858486121766\n",
"total variance 0.3926204053521799 0.3411904320177375 0.21595534394735755 0.1667146015125302\n",
"0.0018922675765916165 0.27346756436567854 0.12782978158639682\n",
"decay 0.8965735036758975\n"
]
}
],
"source": [
"weight = 'weights_net1/weights2/rnn_1515tanh512_checkpoint39_0_9'\n",
"Info_A_pre, Info_I_pre, Info_P_pre, Info_A_pre_std, Info_I_pre_std, Info_P_pre_std = Memory(weight, k_action = 1, k_stim = 1, k_internal = 1.)\n",
"print ('decay', Info_I_pre[-1]/Info_I_pre[0])"
]
},
{
"cell_type": "code",
"execution_count": 28,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"total variance 0.3890941583610254 0.38792269580285177 0.1565308027528738 0.15878011673607562\n",
"total variance 0.3890941583610254 0.006038961682655473 0.1565308027528738 -1.4909264540848963\n",
"total variance 0.3890941583610254 0.3889066171269333 0.1565308027528738 0.156397185146751\n",
"total variance 0.3890941583610254 0.38890821586223917 0.1565308027528738 0.15627826525581323\n",
"total variance 0.07645866431985703 0.07604381960101966 0.026740480678015703 0.027197663553027624\n",
"total variance 0.07645866431985703 0.00205811561424324 0.026740480678015703 -0.1316365903117495\n",
"total variance 0.07645866431985703 0.07588825701660207 0.026740480678015703 0.026583619423885502\n",
"total variance 0.07645866431985703 0.07586605581198055 0.026740480678015703 0.026183714513602525\n",
"0.0007931536385055282 0.22872787269199185 0.0007682497720048599\n"
]
}
],
"source": [
"weight = 'weights_net1/weights2/rnn_1515tanh512_checkpoint0_0_9'\n",
"Info_A_echo, Info_I_echo, Info_P_echo, Info_A_echo_std, Info_I_echo_std, Info_P_echo_std = Memory(weight, k_action = 1, k_stim = 1, k_internal = 1.)\n"
]
},
{
"cell_type": "code",
"execution_count": 34,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"total variance 0.37142232931385033 0.36997640590993425 0.15054802103090953 0.15328924457735915\n",
"total variance 0.37142232931385033 0.003745210031667177 0.15054802103090953 -1.2068524604190993\n",
"total variance 0.37142232931385033 0.3712471560423465 0.15054802103090953 0.15049521292227086\n",
"total variance 0.37142232931385033 0.37130669775081776 0.15054802103090953 0.15073991319265462\n",
"total variance 0.5585121331631487 0.5583796730887401 0.2430539836013729 0.24348393319219835\n",
"total variance 0.5585121331631487 0.0023851624742537757 0.2430539836013729 -3.5099734656466763\n",
"total variance 0.5585121331631487 0.5583625965896789 0.2430539836013729 0.24281620388328276\n",
"total variance 0.5585121331631487 0.5583799366009141 0.2430539836013729 0.24291266396094408\n",
"0.0007891917391623648 0.4619020449855391 0.00028626898512048026\n"
]
}
],
"source": [
"weight = 'weights_cpu8/rnn_1515tanh512_checkpoint0'\n",
"\n",
"Info_A_pre2, Info_I_pre2, Info_P_pre2, Info_A_pre2_std, Info_I_pre2_std, Info_P_pre2_std = Memory(weight, k_action = 1, k_stim = 1, k_internal = 1.)\n"
]
},
{
"cell_type": "code",
"execution_count": 62,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"total variance 0.35743713650016773 0.35556522930974654 0.14829464042823223 0.15222713251692893\n",
"total variance 0.35743713650016773 0.0068898356969406965 0.14829464042823223 -1.0882732860562254\n",
"total variance 0.35743713650016773 0.3572777924528901 0.14829464042823223 0.1480290901282575\n",
"total variance 0.35743713650016773 0.3572620463351035 0.14829464042823223 0.14819110304922256\n",
"total variance 0.0633601570902253 0.06297596392993922 0.024553286557785 0.02502470822115068\n",
"total variance 0.0633601570902253 0.0019464516322192983 0.024553286557785 -0.09387908959012267\n",
"total variance 0.0633601570902253 0.06280443530717227 0.024553286557785 0.023885982614831894\n",
"total variance 0.0633601570902253 0.06296725765432704 0.024553286557785 0.024354889372640475\n",
"0.0011280501753536385 0.20598050313061653 0.000641527715646583\n"
]
}
],
"source": [
"weight = 'weights_cpu8/rnn_1515tanh512_checkpoint0'\n",
"\n",
"Info_A_pre2_s, Info_I_pre2_s, Info_P_pre2_s, Info_A_pre2_std_s, Info_I_pre2_std_s, Info_P_pre2_std_s = Memory(weight, shuffle = True)\n"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {},
"outputs": [],
"source": [
"np.save('info_A_shuffle', Info_A)\n",
"np.save('info_I_shuffle', Info_I)\n",
"np.save('info_p_shuffle', Info_P)\n",
"np.save('info_A_shuffle_std', Info_A_std)\n",
"np.save('info_I_shuffle_std', Info_I_std)\n",
"np.save('info_p_shuffle_std', Info_P_std)"
]
},
{
"cell_type": "code",
"execution_count": 25,
"metadata": {},
"outputs": [],
"source": [
"np.save('info_A_pre', Info_A_pre)\n",
"np.save('info_I_pre', Info_I_pre)\n",
"np.save('info_p_pre', Info_P_pre)\n",
"np.save('info_A_pre_std', Info_A_pre_std)\n",
"np.save('info_I_pre_std', Info_I_pre_std)\n",
"np.save('info_p_pre_std', Info_P_pre_std)"
]
},
{
"cell_type": "code",
"execution_count": 35,
"metadata": {},
"outputs": [],
"source": [
"np.save('info_A_pre2', Info_A_pre2)\n",
"np.save('info_I_pre2', Info_I_pre2)\n",
"np.save('info_p_pre2', Info_P_pre2)\n",
"np.save('info_A_pre2_std', Info_A_pre2_std)\n",
"np.save('info_I_pre2_std', Info_I_pre2_std)\n",
"np.save('info_p_pre2_std', Info_P_pre2_std)"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {},
"outputs": [],
"source": [
"np.save('info_A_echo', Info_A_echo)\n",
"np.save('info_I_echo', Info_I_echo)\n",
"np.save('info_p_echo', Info_P_echo)\n",
"np.save('info_A_echo_std', Info_A_echo_std)\n",
"np.save('info_I_echo_std', Info_I_echo_std)\n",
"np.save('info_p_echo_std', Info_P_echo_std)"
]
},
{
"cell_type": "code",
"execution_count": 64,
"metadata": {},
"outputs": [],
"source": [
"np.save('info_A_pre2_s', Info_A_pre2_s)\n",
"np.save('info_I_pre2_s', Info_I_pre2_s)\n",
"np.save('info_p_pre2_s', Info_P_pre2_s)\n",
"np.save('info_A_pre2_std_s', Info_A_pre2_std_s)\n",
"np.save('info_I_pre2_std_s', Info_I_pre2_std_s)\n",
"np.save('info_p_pre2_std_s', Info_P_pre2_std_s)"
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(0.293569982016617,\n",
" 0.22370647148053782,\n",
" 0.013565131747016146,\n",
" 0.2883806652741374,\n",
" 0.8965735036758975,\n",
" 0.14995934945114472)"
]
},
"execution_count": 27,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"Info_I[0], Info_I[1]/Info_I[0], Info_P[0], Info_I_pre[0], Info_I_pre[1]/Info_I_pre[0], Info_P_pre[0]"
]
},
{
"cell_type": "code",
"execution_count": 69,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"([,\n",
" ,\n",
" ],\n",
" )"
]
},
"execution_count": 69,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAEBCAYAAACQbKXWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"ax = subplot(1, 1, 1)\n",
"ax.spines['top'].set_visible(False)\n",
"ax.spines['right'].set_visible(False)\n",
"set_stim = [Info_I_echo[0], Info_I_pre[0], Info_I[0], Info_I_pre2[0], Info_I_pre2_s[0]]\n",
"set_stim_std = [Info_I_echo_std[0], Info_I_pre_std[0], Info_I_std[0], Info_I_pre2_std[0], Info_I_pre2_std_s[0]]\n",
"barlist = plt.bar(np.arange(5), height = set_stim, yerr = set_stim_std)\n",
"barlist[0].set_color('b')\n",
"barlist[1].set_color('r')\n",
"barlist[2].set_color('m')\n",
"barlist[3].set_color('g')\n",
"barlist[4].set_color('y')\n",
"plt.xticks([0, 1, 2, 3, 4], ['echo', 'PosNet', 'NP', 'MemNet', 'NM'], size = 15)\n",
"plt.yticks([0, 0.2, 0.4], size = 15)"
]
},
{
"cell_type": "code",
"execution_count": 74,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"([,\n",
" ,\n",
" ],\n",
" )"
]
},
"execution_count": 74,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAEDCAYAAADdpATdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"ax = subplot(1, 1, 1)\n",
"ax.spines['top'].set_visible(False)\n",
"ax.spines['right'].set_visible(False)\n",
"set_path = [Info_P_echo[0], Info_P_pre[0], Info_P[0], Info_P_pre2[0], Info_P_pre2_s[0]]\n",
"set_path_std = [Info_P_echo_std[0], Info_P_pre_std[0], Info_P_std[0], Info_P_pre2_std[0], Info_P_pre2_std_s[0]]\n",
"barlist = plt.bar(np.arange(5), height = set_path, yerr = set_path_std)\n",
"barlist[0].set_color('b')\n",
"barlist[1].set_color('r')\n",
"barlist[2].set_color('m')\n",
"barlist[3].set_color('g')\n",
"barlist[4].set_color('y')\n",
"plt.xticks([0, 1, 2, 3, 4], ['echo', 'PosNet', 'NP', 'MemNet', 'NM'], size = 15)\n",
"plt.yticks([0, 0.1, 0.2], size = 15)"
]
},
{
"cell_type": "code",
"execution_count": 81,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"([,\n",
" ],\n",
" )"
]
},
"execution_count": 81,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAXEAAAEBCAYAAACOpZVlAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"ax = subplot(1, 1, 1)\n",
"ax.spines['top'].set_visible(False)\n",
"ax.spines['right'].set_visible(False)\n",
"set_stim = [Info_I_echo[1]/Info_I_echo[0], Info_I_pre[1]/Info_I_pre[0], Info_I[1]/Info_I[0], Info_I_pre2[1]/Info_I_pre2[0], \\\n",
" Info_I_pre2_s[1]/Info_I_pre2_s[0]]\n",
"set_stim_std = [Info_I_echo_std[1]/Info_I_echo[0], Info_I_pre_std[1]/Info_I_pre[0], Info_I_std[1]/Info_I[0], \\\n",
" Info_I_pre2_std[1]/Info_I_pre2[0], Info_I_pre2_std_s[1]/Info_I_pre2_s[0]]\n",
"barlist = plt.bar(np.arange(5), height = set_stim, yerr = set_stim_std)\n",
"barlist[0].set_color('b')\n",
"barlist[1].set_color('r')\n",
"barlist[2].set_color('m')\n",
"barlist[3].set_color('g')\n",
"barlist[4].set_color('y')\n",
"plt.xticks([0, 1, 2, 3, 4], ['echo', 'PosNet', 'NP', 'MemNet', 'NM'], size = 15)\n",
"plt.yticks([0, 1], size = 15)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Behaviour "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"torch.manual_seed(np.random.randint(1e5))\n",
"hidden = torch.randn(1, 512)\n",
"size = 10\n",
"PC_traces = []\n",
"start = (np.random.randint(2, size +2), np.random.randint(2, size+2))\n",
"for i in range(1):\n",
" weight ='weights_net1_shuffle/rnn_1515tanh512_checkpoint39_2_9'\n",
" game = ValueMaxGame(grid_size = (size, size), holes = 0, random_seed = 4 , set_reward = [(0.5, 0.25), (0.5, 0.75)], input_type = 0, action_control = 1, discount = 0.9, alpha = 1, time_limit=8,\n",
" lam = 0)\n",
" game.net.load_state_dict(torch.load(weight))\n",
" grid = game.grid.grid.copy()\n",
" Pos1, hidden1, dH1, Action, state1, reward1 = trajectory(game, start, reward_control = 0, size = size, test = size//10-1, init_hidden = False, hidden = hidden)\n",
" Pos2, hidden2, dH2, Action, state2, reward2 = trajectory(game, start, reward_control = 1, size = size, test = size//10-1, init_hidden = False, hidden = hidden)\n",
" \n",
" if i == 0:\n",
" plt.matshow(grid)\n",
"# y1 = np.array([p[0] for p in Pos1])\n",
"# x1 = np.array([p[1] for p in Pos1])\n",
" y2 = np.array([p[0] for p in Pos2])\n",
" x2 = np.array([p[1] for p in Pos2])\n",
"\n",
" plt.plot(start[1], start[0], 'rs')\n",
"# plt.quiver(x1[:-1], y1[:-1], x1[1:]-x1[:-1], y1[1:]-y1[:-1], scale_units='xy', angles='xy', scale=1, color = 'm')\n",
" plt.quiver(x2[:-1], y2[:-1], x2[1:]-x2[:-1], y2[1:]-y2[:-1], scale_units='xy', angles='xy', scale=1, color = 'm')\n",
" # plt.quiver(x3[:-1], y3[:-1], x3[1:]-x3[:-1], y3[1:]-y3[:-1], scale_units='xy', angles='xy', scale=1, color = 'r')\n",
" # plt.quiver(x4[:-1], y4[:-1], x4[1:]-x4[:-1], y4[1:]-y4[:-1], scale_units='xy', angles='xy', scale=1, color = 'k')\n",
"\n",
"\n",
"size = 90\n",
"PC_traces = []\n",
"start = (np.random.randint(2, size +2), np.random.randint(2, size+2))\n",
"for i in range(1):\n",
" weight = 'weights_net1_shuffle/rnn_1515tanh512_checkpoint39_2_9'\n",
" game = ValueMaxGame(grid_size = (size, size), holes = 0, random_seed = 4 , set_reward = [(0.5, 0.25), (0.5, 0.75)], input_type = 0, action_control = 1, discount = 0.9, alpha = 1, time_limit=8,\n",
" lam = 0)\n",
" game.net.load_state_dict(torch.load(weight))\n",
" grid = game.grid.grid.copy()\n",
" Pos1, hidden1, dH1, Action, state1, reward1 = trajectory(game, start, reward_control = 0, size = size, test = size//10-1, init_hidden = False, hidden = hidden)\n",
" Pos2, hidden2, dH2, Action, state2, reward2 = trajectory(game, start, reward_control = 1, size = size, test = size//10-1, init_hidden = False, hidden = hidden)\n",
" print (reward1, reward2)\n",
" if i == 0:\n",
" plt.matshow(grid)\n",
" y1 = np.array([p[0] for p in Pos1])\n",
" x1 = np.array([p[1] for p in Pos1])\n",
" y2 = np.array([p[0] for p in Pos2])\n",
" x2 = np.array([p[1] for p in Pos2])\n",
"\n",
" plt.plot(start[1], start[0], 'rs')\n",
" plt.quiver(x1[:-1], y1[:-1], x1[1:]-x1[:-1], y1[1:]-y1[:-1], scale_units='xy', angles='xy', scale=1, color = 'c')\n",
" plt.quiver(x2[:-1], y2[:-1], x2[1:]-x2[:-1], y2[1:]-y2[:-1], scale_units='xy', angles='xy', scale=1, color = 'm')\n",
" # plt.quiver(x3[:-1], y3[:-1], x3[1:]-x3[:-1], y3[1:]-y3[:-1], scale_units='xy', angles='xy', scale=1, color = 'r')\n",
" # plt.quiver(x4[:-1], y4[:-1], x4[1:]-x4[:-1], y4[1:]-y4[:-1], scale_units='xy', angles='xy', scale=1, color = 'k')\n",
"\n",
" \n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Conclusion\n",
"**Internal dynamics rather than decoding performance influences more the performance in uncertain enviroments**\n",
"\n",
"**Consider to think in global loops rather than separated system, thus the coupled dynamical system. This system, which has its genome on its internal weights, will have quite different different pheno-types from the global loops, richness of internal dynamics turns into extenable phenotypes which supports behaviours across scales. The different types of attractors, fix point, limit cycle becomes substrate of strategy in different size rooms. Thus generalization can be done until certain size, which offers complicate behaviour with a \"simple mind\"**\n",
"\n",
"**The triangle between decoding, dynamics and navigation should be like: 1, dynamics coded by internal weight is most fundamental, it gives rise to different representations(phenotypes). 2, The relation between decoding and navigation are reciprocal. not random behaviour gives better enviromental prediction(complement the free energy principle) 2, we should distinguish internal dynamics and global dynamics**"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"anaconda-cloud": {},
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.3"
},
"latex_envs": {
"LaTeX_envs_menu_present": true,
"autoclose": false,
"autocomplete": true,
"bibliofile": "biblio.bib",
"cite_by": "apalike",
"current_citInitial": 1,
"eqLabelWithNumbers": true,
"eqNumInitial": 1,
"hotkeys": {
"equation": "Ctrl-E",
"itemize": "Ctrl-I"
},
"labels_anchors": false,
"latex_user_defs": false,
"report_style_numbering": false,
"user_envs_cfg": false
},
"toc": {
"nav_menu": {},
"number_sections": true,
"sideBar": true,
"skip_h1_title": false,
"title_cell": "Table of Contents",
"title_sidebar": "Contents",
"toc_cell": true,
"toc_position": {
"height": "759px",
"left": "0px",
"right": "1228px",
"top": "67px",
"width": "212px"
},
"toc_section_display": "block",
"toc_window_display": true
},
"varInspector": {
"cols": {
"lenName": 16,
"lenType": 16,
"lenVar": 40
},
"kernels_config": {
"python": {
"delete_cmd_postfix": "",
"delete_cmd_prefix": "del ",
"library": "var_list.py",
"varRefreshCmd": "print(var_dic_list())"
},
"r": {
"delete_cmd_postfix": ") ",
"delete_cmd_prefix": "rm(",
"library": "var_list.r",
"varRefreshCmd": "cat(var_dic_list()) "
}
},
"types_to_exclude": [
"module",
"function",
"builtin_function_or_method",
"instance",
"_Feature"
],
"window_display": false
}
},
"nbformat": 4,
"nbformat_minor": 2
}