NN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "def b_dense(x_inp,w,b,activation):\n",
    "    x_inp=np.array(x_inp).reshape(1,-1)\n",
    "    w=np.array(w).T\n",
    "    b=np.array(b).reshape(1,-1)\n",
    "    Z=np.matmul(x_inp,w)+b\n",
    "\n",
    "    return activation(Z),Z"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import math\n",
    "def b_sigmoid(x_inp,diff=False):\n",
    "    def singleSigmoid(z1):\n",
    "        try:\n",
    "            return (1/(1+math.exp(-z1)))\n",
    "        except:\n",
    "            print(z1)\n",
    "    def derivative(z1):\n",
    "        temp1=singleSigmoid(z1)\n",
    "        return (temp1*(1-temp1))\n",
    "    if diff:\n",
    "        return np.vectorize(derivative)(x_inp)\n",
    "    else:\n",
    "        return np.vectorize(singleSigmoid)(x_inp) #https://stackoverflow.com/a/62865571\n",
    "def b_linear(x_inp,diff=False):\n",
    "    if diff:\n",
    "        return np.vectorize((lambda a:1))(x_inp)\n",
    "    else:\n",
    "        return x_inp\n",
    "def b_threshold(x_inp):\n",
    "    return np.vectorize((lambda a:1 if a>0 else 0))(x_inp)\n",
    "def b_relu(x_inp,diff=False):\n",
    "    if diff:\n",
    "        return np.vectorize((lambda a:1 if a>0 else 0.01))(x_inp)\n",
    "    else:\n",
    "        return np.vectorize((lambda a:a if a>0 else 0.01*a))(x_inp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def b_sequential(x_inp,Weights,Biases,Activation):\n",
    "    f_x=x_inp\n",
    "    Zs=[]\n",
    "    for i in range(len(Activation)):\n",
    "        f_x,z_x=(b_dense)(f_x,Weights[i],Biases[i],Activation[i])\n",
    "        #print(f'Layer {i+1}: {f_x}')\n",
    "        Zs.append(z_x)\n",
    "    return f_x,Zs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1, 2, 3]])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array([1,2,3]).reshape(1,-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def XOR(x_inp):\n",
    "    output,layers=b_sequential(x_inp,[[[1,1],[-1,-1]],[[1,1]]],[[-0.5,1.5],[-1.5]],[b_threshold for i in range(2)])\n",
    "    return output,layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([[0.5, 0.5]]), array([[0.5]])]\n",
      "[[1]]\n"
     ]
    }
   ],
   "source": [
    "output0,layers0=XOR([1,0])\n",
    "print(layers0)\n",
    "print(output0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.75136507 0.77292847]]\n"
     ]
    }
   ],
   "source": [
    "def Vid(x_inp):\n",
    "    output=b_sequential(x_inp,[[[0.15,0.2],[0.25,0.3]],[[0.4,0.45],[0.5,0.55]]],[[0.35,0.35],[0.6,0.6]],[b_sigmoid for i in range(2)])\n",
    "    return output\n",
    "o1,k1=(Vid([[0.05,0.1]]))\n",
    "print(o1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def b_error(output,target):\n",
    "    return 0.5*sum([(target[i]-output[i])**2 for i in range(len(target))])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def updateVal(a,b,step):\n",
    "    return [list(ai + step * bi for ai, bi in zip(sub_a, sub_b)) for sub_a, sub_b in zip(a, b)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cost at #0: 0.2983711087600027\n",
      "Cost at #1000: 2.0254878410674998e-05\n",
      "Cost at #2000: 1.303401740146918e-055\n",
      "Cost at #3000: 8.556149642287105e-065\n",
      "Cost at #4000: 4.959091133189774e-066\n",
      "Cost at #5000: 2.431847837573228e-066\n",
      "Cost at #6000: 1.0029589201149795e-06\n",
      "Cost at #7000: 3.571911381942394e-077\n",
      "([array([[0.15101984, 0.20203968],\n",
      "       [0.24992326, 0.29984652]]), array([[-0.69193635, -0.6393511 ],\n",
      "       [ 1.13089895,  1.18137643]])], [array([0.37039675, 0.34846523]), array([-3.88816526,  3.20318898])])\n"
     ]
    }
   ],
   "source": [
    "def b_backprop(x_inps,Weights,Biases,Activation,Targets,setLdAlNum):\n",
    "    changeB=[np.zeros(len(Biases[b00])) for b00 in range(len(Biases))]\n",
    "    changeW=[np.array(w00)*0 for w00 in Weights]\n",
    "    lambda_,alpha_,numIterations=setLdAlNum\n",
    "    cost=0\n",
    "\n",
    "    BiasesOutput=[np.array(b00) for b00 in Biases]\n",
    "    WeightsOutput=[np.array(w00) for w00 in Weights]\n",
    "\n",
    "    TrainSize=len(Targets)\n",
    "    for gradientDescent in range(numIterations):\n",
    "        \n",
    "        cost=0\n",
    "        delCdelB=[np.zeros(len(Biases[b00])) for b00 in range(len(Biases))]\n",
    "        delCdelW=[np.array(w00)*0 for w00 in Weights]\n",
    "        for i in range(TrainSize):\n",
    "            Weights=WeightsOutput\n",
    "            Biases=BiasesOutput\n",
    "\n",
    "            outputi,layersi=b_sequential(x_inps[i],Weights,Biases,Activation)\n",
    "            cost+= b_error(outputi[0],Targets[i])\n",
    "            delCdelzcurr=[((outputi[0]-Targets[i])*Activation[-1](layersi[-1][0],diff=True))]\n",
    "            for ll in range(len(Biases)-1,0,-1):\n",
    "                nextLayers=np.array(delCdelzcurr[0])\n",
    "                outWeights=np.array(Weights[ll])\n",
    "                delCdelzcurr.insert(0,np.matmul(nextLayers,outWeights)*Activation[ll](layersi[ll-1][0],diff=True))\n",
    "            #print(delCdelzcurr)\n",
    "\n",
    "            delCdelwcurr=[]\n",
    "            for j in range(len(Weights)):\n",
    "                if j==0:\n",
    "                    delCdelwcurr.append(np.matmul(np.array([delCdelzcurr[j]]).T,np.array([x_inps[i]])))\n",
    "                else:\n",
    "                    delCdelwcurr.append(np.matmul(np.array([delCdelzcurr[j]]).T,np.array([Activation[j-1](layersi[j-1][0],diff=True)])))\n",
    "\n",
    "            #print(delCdelwcurr)\n",
    "\n",
    "            delCdelB=[delCdelB[b00]+delCdelzcurr[b00] for b00 in range(len(Biases))]\n",
    "            delCdelW=[delCdelW[w00]+delCdelwcurr[w00] for w00 in range(len(Weights))]\n",
    "            #\n",
    "        cost=cost/TrainSize\n",
    "        print(f\"Cost at #{gradientDescent}: {cost}\\r\",end='')\n",
    "        if gradientDescent%1000==0:\n",
    "            print(f\"Cost at #{gradientDescent}: {cost}\")\n",
    "\n",
    "\n",
    "        BiasesOutput=[BiasesOutput[b00]-(lambda_/TrainSize)*delCdelB[b00]+(alpha_)*changeB[b00] for b00 in range(len(Biases))]\n",
    "        WeightsOutput=[WeightsOutput[b00]-(lambda_/TrainSize)*delCdelW[b00]+(alpha_)*changeW[b00] for b00 in range(len(Biases))]\n",
    "        #https://youtu.be/iudXf5n_3ro\n",
    "        changeB=[-(lambda_/TrainSize)*delCdelB[b00]+(alpha_)*changeB[b00] for b00 in range(len(Biases))]\n",
    "        changeW=[-(lambda_/TrainSize)*delCdelW[b00]+(alpha_)*changeW[b00] for b00 in range(len(Biases))]\n",
    "\n",
    "\n",
    "    print(f\"Cost at #{numIterations}: {cost}\")\n",
    "    return WeightsOutput,BiasesOutput\n",
    "            \n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "print(b_backprop([[0.05,0.1]],[[[0.15,0.2],[0.25,0.3]],[[0.4,0.45],[0.5,0.55]]],[[0.35,0.35],[0.6,0.6]],[b_sigmoid for i in range(2)],[[0.01, 0.99]],(0.5,0.9,7000)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a=([[1],[2,3]])\n",
    "updateVal(0*a,a,1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0, 0],\n",
       "       [0, 0],\n",
       "       [0, 0]])"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "abc=np.array([[1,2],[3,4],[5,6]])*0\n",
    "abc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "def b_weightsbiasesrandom(layerinfo):\n",
    "    brandom=[[random.uniform(-1, 1) for _ in range(num0)] for num0 in layerinfo[1:]]\n",
    "    wrandom=[np.array([random.uniform(-1, 1) for _ in range(layerinfo[i]*layerinfo[i+1])]).reshape(layerinfo[i+1],layerinfo[i]) for i in range(len(layerinfo)-1)]\n",
    "    return wrandom,brandom"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cost at #0: 0.15145803977753974\n",
      "Cost at #1000: 0.12862714034737827\n",
      "Cost at #2000: 0.12853700999166195\n",
      "Cost at #3000: 0.13131219550012854\n",
      "Cost at #4000: 0.13321338424294274\n",
      "Cost at #5000: 0.13336098109870262\n",
      "Cost at #6000: 0.13118294647697637\n",
      "Cost at #7000: 0.12590332905497842\n",
      "Cost at #8000: 0.11629879976873098\n",
      "Cost at #9000: 0.10225977258582361\n",
      "Cost at #10000: 0.08811925322072556\n",
      "Cost at #11000: 0.08048486386404002\n",
      "Cost at #12000: 0.07586419491074733\n",
      "Cost at #13000: 0.07257533835414354\n",
      "Cost at #14000: 0.06992548813175231\n",
      "Cost at #15000: 0.06755244517172491\n",
      "Cost at #16000: 0.06526880468921889\n",
      "Cost at #17000: 0.06298972884606432\n",
      "Cost at #18000: 0.060688276571714875\n",
      "Cost at #19000: 0.058367058833339736\n",
      "Cost at #20000: 0.056041187750471046\n",
      "Cost at #21000: 0.053728936580046086\n",
      "Cost at #22000: 0.051447299867484746\n",
      "Cost at #23000: 0.048904750723986994\n",
      "Cost at #24000: 0.044015865794573644\n",
      "Cost at #25000: 0.040545965896328175\n",
      "Cost at #26000: 0.038049768205540103\n",
      "Cost at #27000: 0.036051387955033085\n",
      "Cost at #28000: 0.034304931962540364\n",
      "Cost at #29000: 0.032689058321053664\n",
      "Cost at #30000: 0.031151185049453037\n"
     ]
    }
   ],
   "source": [
    "xorw,xorb=b_weightsbiasesrandom((2,2,1))\n",
    "xoractivation=[b_relu,b_sigmoid]\n",
    "xorw,xorb=b_backprop([[0,0],[0,1],[1,0],[1,1]],xorw,xorb,xoractivation,[[0],[1],[1],[0]],(0.01,0.05,30000))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.22643772]]\n",
      "[[0.7846937]]\n",
      "[[0.69596901]]\n",
      "[[0.24317014]]\n"
     ]
    }
   ],
   "source": [
    "for inp in [[0,0],[0,1],[1,0],[1,1]]:\n",
    "    out00,layer00=b_sequential(inp,xorw,xorb,xoractivation)\n",
    "    print(out00)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cost at #0: 176.86227762935516\n",
      "Cost at #1000: 5.003295992740267\n",
      "Cost at #2000: 4.3390065028951415\n",
      "Cost at #3000: 3.7634651351668938\n",
      "Cost at #4000: 3.2648152678571895\n",
      "Cost at #5000: 2.8327843069039463\n",
      "Cost at #6000: 2.4584720620632944\n",
      "Cost at #7000: 2.1341673957298846\n",
      "Cost at #8000: 1.8531893672114736\n",
      "Cost at #9000: 1.6097495999004277\n",
      "Cost at #10000: 1.3988330359934653\n",
      "Cost at #11000: 1.2160946222096451\n",
      "Cost at #12000: 1.0577697981481522\n",
      "Cost at #13000: 0.9205969432729031\n",
      "Cost at #14000: 0.8017501848686711\n",
      "Cost at #15000: 0.6987811827578861\n",
      "Cost at #16000: 0.6095686914955692\n",
      "Cost at #17000: 0.5322748609827178\n",
      "Cost at #18000: 0.46530737525493165\n",
      "Cost at #19000: 0.40728664947455184\n",
      "Cost at #20000: 0.35701740935754675\n",
      "Cost at #21000: 0.31346406754816997\n",
      "Cost at #22000: 0.27572938967460325\n",
      "Cost at #23000: 0.24303601058892707\n",
      "Cost at #24000: 0.21471042001091496\n",
      "Cost at #25000: 0.19016908766682095\n",
      "Cost at #26000: 0.16890644208968558\n",
      "Cost at #27000: 0.15048445543468348\n",
      "Cost at #28000: 0.13452361974817695\n",
      "Cost at #29000: 0.12069512879427166\n",
      "Cost at #30000: 0.10871410437811832\n",
      "Cost at #31000: 0.09833372762269722\n",
      "Cost at #32000: 0.08934015429859848\n",
      "Cost at #33000: 0.08154810945844752\n",
      "Cost at #34000: 0.07479707062183061\n",
      "Cost at #35000: 0.06894796088125743\n",
      "Cost at #36000: 0.06388028380440637\n",
      "Cost at #37000: 0.059489641109311286\n",
      "Cost at #38000: 0.055685581974595756\n",
      "Cost at #39000: 0.052389739678756056\n",
      "Cost at #40000: 0.049534217181770145\n",
      "Cost at #41000: 0.047060188390681774\n",
      "Cost at #42000: 0.044916686294097416\n",
      "Cost at #43000: 0.043059553000171724\n",
      "Cost at #44000: 0.041450530048011186\n",
      "Cost at #45000: 0.040056470252171055\n",
      "Cost at #46000: 0.038848654843589514\n",
      "Cost at #47000: 0.037802201839523676\n",
      "Cost at #48000: 0.036895553454399065\n",
      "Cost at #49000: 0.036110031991843906\n",
      "Cost at #50000: 0.035430088057462776\n"
     ]
    }
   ],
   "source": [
    "lineareg_w,lineareg_b=b_weightsbiasesrandom((1,1))\n",
    "\n",
    "linear_activation=[b_linear]\n",
    "lineareg_inp=[[xinp] for xinp in [3.6,11.63,14,17.57,20.54,23,25.4,26.7]]\n",
    "lineareg_expectedoutput=[[yinp] for yinp in [9.85,13.4,14,15.4,16.4,18,18.87,19.8]]\n",
    "lineareg_w,lineareg_b=b_backprop(lineareg_inp,lineareg_w,lineareg_b,linear_activation,lineareg_expectedoutput,(0.0005,0,50000))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([[0.43062433]])] [array([8.01219617])]\n"
     ]
    }
   ],
   "source": [
    "print(lineareg_w,lineareg_b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From c:\\Python311\\Lib\\site-packages\\keras\\src\\losses.py:2976: The name tf.losses.sparse_softmax_cross_entropy is deprecated. Please use tf.compat.v1.losses.sparse_softmax_cross_entropy instead.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras import datasets\n",
    "#https://www.kaggle.com/code/abdelrahmanramadan2/mnist?scriptVersionId=152380489&cellId=2\n",
    "(x_train, y_train), (x_test, y_test) = datasets.mnist.load_data()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "60000"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(x_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 100 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.figure(figsize=(6, 6))\n",
    "for i in range(100):\n",
    "  plt.subplot(10,10, i + 1)\n",
    "  plt.imshow(x_train[i], cmap='gray')\n",
    "  plt.axis('off')\n",
    "  plt.title('Digit {}'.format(y_train[i]))\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def flatten2d(thelist):\n",
    "    return ([xyz/255 for yzx in thelist for xyz in yzx])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1, 1, ..., 1, 0, 1], dtype=uint8)"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(x_train[1]).shape\n",
    "28**2\n",
    "\n",
    "digit01=[i for i in range(0,60000) if y_train[i]==0 or y_train[i]==1]\n",
    "x_train01=x_train[digit01]\n",
    "y_train01=y_train[digit01]\n",
    "y_train01"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cost at #0: 0.22601915573661713\n",
      "Cost at #100: 0.2259498259036038\n"
     ]
    }
   ],
   "source": [
    "digit_w,digit_b=b_weightsbiasesrandom((784,800,1))\n",
    "\n",
    "digit_activation=[b_relu,b_sigmoid]\n",
    "digit_inp=[flatten2d(x_train01[i]) for i in range(100)]\n",
    "#digit_expectedoutput=[[int(yinp==r10) for r10 in range(10)] for yinp in y_train01[:100]]\n",
    "digit_expectedoutput=[[yinp] for yinp in y_train01[:100]]\n",
    "digit_w,digit_b=b_backprop(digit_inp,digit_w,digit_b,digit_activation,digit_expectedoutput,(0.00005,0,100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "for k01 in range(100):\n",
    "    digit_inp=[flatten2d(x_train01[i]) for i in range(100*k01,100+100*k01)]\n",
    "    digit_expectedoutput=[[yinp] for yinp in y_train01[:100]]\n",
    "    digit_w,digit_b=b_backprop(digit_inp,digit_w,digit_b,digit_activation,digit_expectedoutput,(0.005,0,100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[132.52452438]]\n",
      "0\n",
      "[array([[-0.67082781,  0.93462591,  0.95419774, ..., -0.66915081,\n",
      "        -0.09683491,  0.8668029 ],\n",
      "       [ 0.83127791, -0.84910539,  0.88732978, ..., -0.70956392,\n",
      "        -0.01441394,  0.90463695],\n",
      "       [-0.47951885,  0.62171697,  0.22550166, ..., -0.65194848,\n",
      "         0.06125151,  0.050418  ],\n",
      "       ...,\n",
      "       [-0.61920903,  0.54814818,  0.19574696, ...,  0.13252281,\n",
      "         0.68583668, -0.87462425],\n",
      "       [-0.23947153, -0.39167674,  0.90604698, ..., -0.51666283,\n",
      "         0.50954737,  0.96857956],\n",
      "       [ 0.81730254, -0.63532028,  0.59225843, ..., -0.50944317,\n",
      "        -0.97953498,  0.66447297]]), array([[-5.45030294e-01,  2.90664953e-01,  3.02407258e-01,\n",
      "        -8.81860861e-01,  2.20880338e-01, -8.56300140e-01,\n",
      "        -5.39143617e-01,  2.64430820e-01, -8.66811273e-01,\n",
      "         6.94510356e-01, -9.91893830e-01,  2.53586356e-02,\n",
      "        -3.89596019e-01, -8.33694711e-01, -4.84249751e-01,\n",
      "        -3.37794748e-01, -7.72108899e-01, -5.83266112e-01,\n",
      "        -8.13851899e-01,  5.21582386e-01, -4.20067812e-01,\n",
      "         8.94115861e-01, -6.68372967e-01,  6.56751702e-01,\n",
      "         2.42163210e-01, -2.20028781e-01,  3.75035958e-01,\n",
      "        -1.41599001e-01,  6.83938716e-02,  9.44287137e-01,\n",
      "        -7.20978290e-01,  6.72775789e-01, -4.49808816e-01,\n",
      "         2.93605883e-02,  2.24186517e-02,  1.29313535e-01,\n",
      "         8.01766787e-01,  6.35683862e-01,  8.59144902e-01,\n",
      "         1.01766249e-01,  2.22822628e-01, -1.58349632e-01,\n",
      "        -9.16391388e-01, -4.64803695e-01,  2.75753857e-01,\n",
      "        -1.02550371e-01, -9.56779217e-01, -6.18250382e-01,\n",
      "        -4.71514004e-01,  4.40283371e-01, -9.82046324e-01,\n",
      "        -3.93586104e-01, -3.13721739e-01, -4.46920257e-01,\n",
      "        -1.62343982e-01,  3.02875562e-01,  3.46008461e-02,\n",
      "         9.43699315e-01, -8.97408777e-01, -5.48036501e-01,\n",
      "        -4.72844313e-02, -7.71596371e-01, -4.33960588e-01,\n",
      "         6.37611699e-01,  1.36936405e-01,  2.59370232e-01,\n",
      "         9.91868542e-01,  3.59380555e-01,  6.34166785e-01,\n",
      "        -8.98721027e-01,  6.30270271e-01,  5.65615409e-02,\n",
      "         7.34932499e-01,  1.64144521e-01,  4.98449775e-01,\n",
      "        -7.48383997e-01,  7.06724534e-01, -4.60265997e-01,\n",
      "         4.69855840e-01, -9.37695664e-01, -2.27880141e-01,\n",
      "        -5.96085617e-01,  3.39436582e-01,  5.30640835e-01,\n",
      "        -3.69946391e-01,  4.71253180e-01, -4.41799138e-01,\n",
      "        -2.97206649e-01, -9.00285162e-01,  2.90247498e-01,\n",
      "        -3.90179576e-01,  1.17488005e-01, -1.73838811e-02,\n",
      "         4.53487722e-01,  6.44869254e-02, -4.16130350e-01,\n",
      "         6.44095517e-02, -6.31525994e-02,  2.12885604e-01,\n",
      "         5.73240012e-01, -4.36082524e-01, -2.24011494e-01,\n",
      "         9.62120420e-01,  7.42748500e-01, -3.23360031e-01,\n",
      "        -3.65993741e-01,  3.30165928e-01, -1.25971055e-01,\n",
      "         8.01132451e-01,  8.43545339e-01,  8.59786389e-01,\n",
      "         1.49712092e-02, -4.64118530e-01,  3.52332274e-01,\n",
      "        -1.51392062e-01,  5.66903408e-01,  2.00705274e-02,\n",
      "        -7.39274785e-01,  5.21417472e-01,  9.02041377e-01,\n",
      "         6.94583908e-01, -2.28630639e-01,  3.62648079e-01,\n",
      "         9.48358191e-01, -6.87525797e-01,  3.34183299e-02,\n",
      "        -5.26540062e-02,  8.98379514e-01, -3.30178876e-01,\n",
      "        -6.32450286e-01,  1.55509236e-01, -9.65204923e-01,\n",
      "         5.35123299e-01,  8.23533436e-01, -2.35995590e-01,\n",
      "        -6.49829143e-01, -4.23638630e-01,  2.95733258e-01,\n",
      "         5.34060108e-01, -8.24059536e-01, -5.48329663e-01,\n",
      "        -2.40233385e-01, -4.87388074e-01, -5.39898255e-01,\n",
      "        -8.15379977e-01, -9.15813679e-01, -1.50881526e-01,\n",
      "        -9.21879536e-01,  5.75814423e-01, -8.44014923e-01,\n",
      "         2.86338222e-01, -9.52096174e-01,  7.79243451e-01,\n",
      "         4.60291271e-01,  6.15057302e-01,  9.27118350e-01,\n",
      "        -5.40464355e-01, -1.08054846e-01, -1.99438818e-01,\n",
      "        -6.25546632e-01,  5.89010882e-01,  1.23303285e-01,\n",
      "         7.63597170e-01, -7.96567365e-01,  1.60330536e-02,\n",
      "        -2.57972829e-01, -2.76591501e-01, -7.61682599e-01,\n",
      "        -7.33740105e-01, -4.70241384e-01,  9.19057490e-01,\n",
      "        -2.31606805e-01,  7.97873801e-01, -9.87153199e-01,\n",
      "        -7.07271878e-01,  2.42680832e-02, -5.74655399e-01,\n",
      "         7.34589527e-01,  6.88429378e-01,  3.80777166e-01,\n",
      "        -5.78738136e-01,  5.57996765e-01,  4.62508896e-01,\n",
      "         7.35065398e-01,  9.28602031e-01,  3.84295492e-01,\n",
      "        -7.01579782e-01,  6.80078287e-01, -6.73738353e-01,\n",
      "         1.86883191e-04, -4.48645915e-01,  1.62270232e-01,\n",
      "         4.21985160e-01,  4.36491459e-01,  2.00315225e-01,\n",
      "        -5.57269605e-02, -5.96964024e-01,  6.85408276e-01,\n",
      "         7.19280368e-01,  6.63093989e-01, -9.01393684e-01,\n",
      "        -6.09915491e-01,  8.86117635e-02,  8.79250190e-01,\n",
      "         2.79633900e-01, -1.43299740e-01, -6.53780503e-01,\n",
      "        -2.71780271e-01,  8.76571659e-01,  1.40464358e-01,\n",
      "         1.14569865e-01, -6.28397817e-01, -9.99424119e-02,\n",
      "        -5.53940541e-01,  8.88940645e-01,  4.40863209e-01,\n",
      "        -7.62079626e-01,  5.75152092e-01, -1.73974351e-01,\n",
      "         4.74971924e-01, -7.27566964e-01,  8.93290439e-01,\n",
      "        -1.75247884e-01, -4.06909992e-02,  6.30789745e-01,\n",
      "         2.19377652e-02,  9.30334760e-01, -1.15955641e-01,\n",
      "         2.88330772e-01, -3.32056205e-01, -7.30848222e-02,\n",
      "         1.49608876e-01, -7.70004941e-01,  8.27781667e-01,\n",
      "         7.68315507e-01, -3.14341630e-01,  6.22161849e-01,\n",
      "         3.17423332e-01,  3.22609310e-02, -9.17956455e-01,\n",
      "         6.60276592e-01, -7.65641605e-01, -8.36151122e-01,\n",
      "        -2.97333134e-01, -8.93285405e-01,  8.94159440e-01,\n",
      "        -3.85101333e-01,  1.83158162e-01,  1.15902629e-01,\n",
      "         4.56511222e-01,  2.42492735e-01, -4.86169325e-01,\n",
      "         2.61244021e-01, -6.54904940e-01,  4.93858688e-02,\n",
      "        -2.16758261e-01, -4.52027733e-01, -1.00672923e-01,\n",
      "        -3.36217447e-01,  8.52710240e-01,  8.72458238e-01,\n",
      "        -5.34434915e-02, -8.33307702e-01,  5.80431628e-01,\n",
      "        -1.16633613e-01, -2.47360172e-01, -4.06223606e-01,\n",
      "        -9.56562320e-01,  9.58900093e-01,  5.97718795e-01,\n",
      "        -9.27345154e-01,  7.93222760e-01, -3.14634699e-01,\n",
      "         5.24330525e-01,  6.25381997e-01, -7.68488986e-01,\n",
      "         7.10768710e-01, -2.05831637e-01,  1.52906193e-01,\n",
      "        -1.49003332e-02,  1.91151425e-01, -1.35864039e-01,\n",
      "        -6.27620576e-01,  9.42912598e-01, -1.93179643e-01,\n",
      "        -6.74426038e-01,  2.28759821e-01,  4.76013156e-01,\n",
      "        -8.61204077e-02,  7.45831819e-01, -1.27782020e-01,\n",
      "        -7.37111591e-01,  7.59623967e-01, -6.93940812e-01,\n",
      "         7.40194128e-01, -7.67801581e-02,  7.41332176e-01,\n",
      "        -3.10260384e-01, -4.18101141e-01, -9.77359211e-01,\n",
      "         2.76096239e-01, -4.95785396e-01,  7.24674757e-02,\n",
      "         9.07534383e-01, -6.88403759e-01,  6.32300310e-01,\n",
      "        -2.87740814e-01, -8.46713086e-01,  8.81472427e-01,\n",
      "         5.86481905e-01,  1.42840762e-01,  1.04537692e-01,\n",
      "         4.63105891e-02,  7.91210799e-01, -4.16510833e-01,\n",
      "        -6.69783838e-02, -8.70277136e-01, -2.87375793e-01,\n",
      "        -8.20067927e-01, -4.29416059e-01, -1.34272711e-01,\n",
      "        -9.28644742e-01,  3.09707958e-01,  8.20349856e-01,\n",
      "        -2.99746945e-01, -1.20158169e-01,  6.99309941e-01,\n",
      "        -9.37209636e-01, -9.18221120e-01, -4.27311194e-01,\n",
      "         9.65041328e-01,  6.27951942e-01, -4.22855674e-01,\n",
      "         3.96013918e-01,  5.05023554e-01,  9.22523848e-01,\n",
      "        -3.93009571e-01, -3.11881458e-01, -4.57030574e-01,\n",
      "        -3.21485738e-01, -2.09986047e-01, -8.53375139e-03,\n",
      "        -2.85140241e-01,  2.27879840e-01,  2.43670274e-01,\n",
      "        -5.03691878e-01, -9.12386712e-01,  8.38261045e-01,\n",
      "        -5.57910708e-01,  3.76921990e-01, -4.44221339e-01,\n",
      "        -5.95756067e-01, -8.21884709e-01,  9.00150197e-01,\n",
      "         7.32948274e-01,  7.13398230e-01, -3.05524225e-01,\n",
      "         8.17036014e-01,  2.18399832e-01,  6.37515914e-01,\n",
      "         5.52859360e-01,  9.47586890e-01,  9.82550333e-01,\n",
      "        -7.67213787e-01, -4.36667277e-01,  9.28509655e-01,\n",
      "        -3.51237629e-01, -7.90760264e-01, -3.41259358e-01,\n",
      "         9.34986418e-01,  8.03611219e-01, -1.17180940e-01,\n",
      "         5.92284333e-01,  1.06412757e-01,  1.00179727e+00,\n",
      "        -9.13024392e-01, -7.94633426e-01, -3.89349333e-02,\n",
      "        -1.48330851e-01, -2.04966300e-01, -9.43606697e-01,\n",
      "        -5.02966308e-01,  4.62755982e-01, -8.32526291e-01,\n",
      "         2.29769260e-02, -2.78930866e-01, -5.44621525e-01,\n",
      "        -1.67242359e-01, -8.53033245e-01, -6.20515385e-01,\n",
      "        -8.31521301e-01, -6.75169135e-01, -3.17948998e-02,\n",
      "        -1.36351814e-01,  3.85531563e-02,  8.90209555e-01,\n",
      "         1.13196607e-01,  8.21642549e-01,  2.98134886e-01,\n",
      "         6.96162748e-01, -1.03752711e-01,  3.54914980e-01,\n",
      "        -4.05338291e-01, -6.57202629e-01,  2.31728389e-01,\n",
      "         7.94521083e-01,  5.47506476e-01,  1.11452387e-01,\n",
      "        -8.88926354e-01, -8.93689268e-01,  6.05253088e-01,\n",
      "         2.57636049e-01, -6.95904226e-01, -8.60467064e-01,\n",
      "        -1.87258348e-01, -1.62925164e-01, -7.18077277e-01,\n",
      "         5.85911322e-01, -8.22355327e-01,  5.18402511e-01,\n",
      "         6.77277936e-02,  5.06621287e-01, -9.75886332e-01,\n",
      "        -9.09308191e-01,  8.72327541e-01, -7.40302447e-01,\n",
      "         6.60094532e-01,  3.24727611e-01,  1.14937669e-01,\n",
      "        -8.20908454e-01, -4.95309163e-01, -6.08843631e-01,\n",
      "        -6.36976780e-01,  2.17435187e-01,  6.40609184e-01,\n",
      "        -7.59494472e-01,  5.93675881e-01,  3.51515645e-01,\n",
      "         1.01548007e-01,  1.12564974e-01, -5.94687408e-02,\n",
      "        -8.69523009e-01,  2.39533234e-01,  2.33624889e-02,\n",
      "        -2.34532094e-01, -8.12358335e-01, -4.03670974e-01,\n",
      "         6.26729049e-01, -9.86791676e-02, -6.05859937e-01,\n",
      "         1.53470936e-01,  4.63777646e-01,  3.77371001e-01,\n",
      "        -2.96545819e-01, -4.98721971e-01,  7.44149065e-01,\n",
      "        -8.13523501e-01, -9.45416111e-01, -5.96222676e-01,\n",
      "        -8.78003368e-01,  4.89473240e-01,  7.77513837e-01,\n",
      "        -4.40612688e-01,  4.63800260e-01,  7.29593571e-01,\n",
      "        -2.13387449e-01, -7.02284936e-02, -3.13933832e-01,\n",
      "        -5.47608520e-01,  5.77688654e-01,  8.04099361e-02,\n",
      "        -5.45136495e-01,  2.35853444e-01,  9.72672859e-01,\n",
      "         6.30101118e-01,  3.28196600e-01, -4.33118321e-02,\n",
      "         9.58715106e-01,  2.98132234e-01,  6.91067001e-01,\n",
      "         8.07690953e-01,  6.96191173e-01,  6.59171037e-01,\n",
      "        -1.90703578e-01,  5.15030462e-01,  9.00332792e-01,\n",
      "        -7.73771360e-01, -6.12671986e-01,  2.15070064e-01,\n",
      "        -8.83772793e-01,  2.91555911e-01,  9.86751806e-01,\n",
      "         7.03732565e-01, -6.83513010e-01, -1.89395366e-01,\n",
      "         1.18167840e-01, -7.26248652e-01,  7.21675468e-02,\n",
      "         3.14753827e-01,  2.64603291e-01,  4.29099185e-01,\n",
      "         7.08132995e-01,  2.15647586e-01, -8.26198666e-01,\n",
      "         6.04669119e-01, -2.85241145e-01, -2.51365501e-01,\n",
      "        -1.21863984e-01,  2.08628569e-01, -1.12698552e-01,\n",
      "         2.27332188e-01, -3.43399915e-01,  6.11808464e-01,\n",
      "         8.60564948e-01, -7.02565265e-01,  7.06025748e-01,\n",
      "        -3.11037142e-02,  9.51144624e-01,  8.16509323e-01,\n",
      "         9.78354100e-01,  1.29745595e-01,  5.37040062e-01,\n",
      "        -3.90866360e-01, -3.31262624e-02,  4.42488373e-01,\n",
      "         3.05281589e-01, -8.48354976e-01, -8.28820266e-01,\n",
      "        -3.82772629e-01, -4.12997567e-01,  8.86401244e-01,\n",
      "        -2.12671965e-02, -3.58301716e-01, -5.86479907e-01,\n",
      "         7.49632615e-01,  8.22494041e-01, -7.24024395e-01,\n",
      "        -1.56852141e-02, -5.21454317e-01,  3.06584849e-01,\n",
      "         4.34072914e-01, -7.58208883e-02, -4.24400801e-03,\n",
      "        -6.10558974e-01,  5.67168060e-02,  6.25417761e-01,\n",
      "        -9.54610594e-01,  7.96336619e-01,  5.37252981e-01,\n",
      "        -3.06818066e-01, -8.88704411e-01,  3.10621932e-01,\n",
      "         7.30110414e-01,  1.71497875e-01, -2.66519923e-01,\n",
      "        -3.35902817e-01,  5.39000289e-01, -5.54592778e-01,\n",
      "         2.05671213e-01,  4.96115052e-01, -8.51561037e-01,\n",
      "        -3.22115177e-01, -3.92333751e-01, -1.13709228e-01,\n",
      "        -8.16393134e-01,  9.20507981e-01,  3.26251547e-01,\n",
      "        -6.76874462e-02, -7.36149526e-01, -2.14002951e-01,\n",
      "         7.20318565e-01,  8.38236250e-01, -3.24540871e-01,\n",
      "        -5.35310840e-01, -9.55756289e-01, -8.13768201e-01,\n",
      "         9.76515102e-01, -9.54941682e-01, -4.56558669e-01,\n",
      "         1.11137235e-01,  2.67550396e-01,  4.37125368e-01,\n",
      "        -2.93672035e-01,  6.86652322e-01,  4.27327290e-01,\n",
      "        -1.74098342e-01,  1.97117209e-01,  9.99013680e-01,\n",
      "         7.23642554e-01, -6.19403109e-01, -9.04625235e-01,\n",
      "        -4.83847977e-01, -4.54827832e-01, -4.72517015e-02,\n",
      "         5.74592463e-01, -8.52584151e-01,  7.84759774e-01,\n",
      "         7.51280264e-01,  6.06764461e-01,  3.02581933e-02,\n",
      "         2.78269902e-01, -1.88596439e-01,  3.26886403e-01,\n",
      "        -3.55496989e-01, -5.36816099e-01,  6.39234693e-01,\n",
      "         2.66103052e-02, -8.83318112e-01, -8.79746731e-01,\n",
      "         3.94308598e-01,  7.36417005e-01,  5.20616550e-01,\n",
      "        -9.46567789e-01, -5.25390225e-01, -6.76919259e-01,\n",
      "         1.39660805e-01,  1.22437696e-01,  9.64461330e-01,\n",
      "        -6.72249479e-01,  3.50638145e-01,  9.53593473e-01,\n",
      "        -6.14768940e-02,  7.39139138e-01, -9.37979631e-01,\n",
      "        -8.28994565e-02,  8.85239744e-01, -9.91708884e-01,\n",
      "        -7.78863593e-01,  6.06411000e-01,  2.15540488e-01,\n",
      "         9.89816730e-01, -8.95034317e-01,  1.61560349e-01,\n",
      "         8.54215186e-01,  8.20366310e-01, -5.35003637e-01,\n",
      "         2.40917650e-01, -4.46500866e-01,  7.85157629e-01,\n",
      "         1.55530352e-01, -2.81790702e-01, -3.42784967e-01,\n",
      "         7.06184024e-01, -2.91423893e-01, -7.10900702e-01,\n",
      "        -6.16592222e-01,  9.84075992e-01,  4.92649659e-01,\n",
      "        -1.66063159e-01,  9.13356855e-01, -3.26747533e-01,\n",
      "        -5.94771075e-01,  1.04763042e-01,  8.74067466e-01,\n",
      "        -8.70947865e-01, -5.25670356e-01,  5.06541774e-01,\n",
      "         3.21878648e-01, -5.54338707e-01, -6.40768701e-01,\n",
      "         8.06761032e-01, -3.93493478e-01, -8.24763610e-01,\n",
      "        -4.70039169e-01,  7.00385191e-01, -7.09694872e-01,\n",
      "         1.72823388e-02, -9.89388980e-01, -3.10340931e-01,\n",
      "         8.46127505e-01, -6.57721561e-01, -2.96556658e-01,\n",
      "        -4.33393955e-01, -8.31579464e-01, -2.70202670e-01,\n",
      "         2.70440897e-01,  1.55195056e-01,  2.54749078e-01,\n",
      "        -9.97304006e-01, -6.90399371e-01, -1.96350704e-01,\n",
      "        -3.70266180e-01, -1.95649471e-01, -6.06431694e-01,\n",
      "         4.85377617e-01, -3.66959009e-01, -7.34444252e-01,\n",
      "        -2.90924730e-01,  5.99215637e-01, -7.72800691e-01,\n",
      "         9.03052731e-01, -6.89345011e-01, -3.16800091e-02,\n",
      "         7.99929951e-01, -4.88118586e-01, -9.16577531e-01,\n",
      "         1.43629693e-02, -6.23948279e-01, -5.40130549e-01,\n",
      "        -6.75195642e-02, -6.11801564e-01,  2.22843712e-01,\n",
      "        -4.08346113e-01,  8.18809347e-01,  2.23496534e-01,\n",
      "        -7.58690271e-01, -7.97720368e-01, -6.69784363e-01,\n",
      "         2.44506892e-01, -3.01346125e-01,  2.96635849e-02,\n",
      "        -4.08241003e-01,  4.58073792e-01, -9.54158944e-01,\n",
      "         5.38378616e-01,  9.79377636e-01,  2.24752519e-02,\n",
      "         1.82794067e-01, -2.95354853e-01, -5.57530559e-01,\n",
      "         4.57071170e-01,  6.40918981e-01,  1.17377254e-01,\n",
      "        -5.57585497e-01,  9.67940613e-01, -6.15313316e-01,\n",
      "         5.62624084e-01, -4.82615801e-01, -9.67097661e-01,\n",
      "         4.14430498e-01,  4.20720971e-01, -5.05575180e-01,\n",
      "        -1.01033963e-01, -8.78227682e-01, -1.71815239e-01,\n",
      "        -5.79129734e-01,  1.18968375e-01,  6.56230525e-01,\n",
      "        -8.90952420e-01,  8.22695581e-01, -9.12883776e-01,\n",
      "        -5.03727480e-01,  9.92474634e-01,  9.02766214e-01,\n",
      "        -8.87177572e-01, -1.38244244e-01, -8.06002357e-01,\n",
      "        -3.37069123e-01, -7.54360251e-01, -8.37676740e-01,\n",
      "        -4.64338177e-04, -3.10306128e-01,  8.27719669e-01,\n",
      "        -5.21415306e-01,  8.35253250e-01,  3.96995456e-02,\n",
      "        -7.76513543e-01,  1.78307578e-01,  7.06301079e-01,\n",
      "         5.51498721e-02,  6.75526109e-01,  6.92500427e-01,\n",
      "         1.40216750e-01, -6.29538300e-01,  3.26000088e-01,\n",
      "         6.07482229e-01,  4.78411460e-01, -9.04400356e-01,\n",
      "         4.35374534e-01,  5.80236247e-01, -4.43801180e-01,\n",
      "         2.70292130e-01,  5.32480630e-01,  3.55572713e-01,\n",
      "         7.31837639e-01,  1.90895103e-01,  7.88123061e-01,\n",
      "         3.35747974e-01, -2.77360133e-01, -8.47019386e-01,\n",
      "         4.48297636e-01,  6.07732278e-01,  1.04173842e-01,\n",
      "        -5.31598520e-01,  2.78362853e-01, -2.92856044e-01,\n",
      "         3.07487761e-01, -1.87266219e-01,  6.00566095e-01,\n",
      "        -8.69475500e-01,  5.64801475e-02, -9.77352401e-01,\n",
      "        -1.30003646e-02,  4.77382186e-01, -4.23306181e-02,\n",
      "         5.31234584e-01,  8.51702322e-01, -9.60990873e-02,\n",
      "         8.63857825e-01,  7.69821294e-01]])]\n"
     ]
    }
   ],
   "source": [
    "digout,diglayer=b_sequential(flatten2d(x_train01[5]),digit_w,digit_b,digit_activation)\n",
    "print(diglayer[1])\n",
    "print(y_train01[5])\n",
    "print(digit_w)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.011764705882352941,\n",
       " 0.07058823529411765,\n",
       " 0.07058823529411765,\n",
       " 0.07058823529411765,\n",
       " 0.49411764705882355,\n",
       " 0.5333333333333333,\n",
       " 0.6862745098039216,\n",
       " 0.10196078431372549,\n",
       " 0.6509803921568628,\n",
       " 1.0,\n",
       " 0.9686274509803922,\n",
       " 0.4980392156862745,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.11764705882352941,\n",
       " 0.1411764705882353,\n",
       " 0.3686274509803922,\n",
       " 0.6039215686274509,\n",
       " 0.6666666666666666,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.8823529411764706,\n",
       " 0.6745098039215687,\n",
       " 0.9921568627450981,\n",
       " 0.9490196078431372,\n",
       " 0.7647058823529411,\n",
       " 0.25098039215686274,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.19215686274509805,\n",
       " 0.9333333333333333,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.984313725490196,\n",
       " 0.36470588235294116,\n",
       " 0.3215686274509804,\n",
       " 0.3215686274509804,\n",
       " 0.2196078431372549,\n",
       " 0.15294117647058825,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.07058823529411765,\n",
       " 0.8588235294117647,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.7764705882352941,\n",
       " 0.7137254901960784,\n",
       " 0.9686274509803922,\n",
       " 0.9450980392156862,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.3137254901960784,\n",
       " 0.611764705882353,\n",
       " 0.4196078431372549,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.803921568627451,\n",
       " 0.043137254901960784,\n",
       " 0.0,\n",
       " 0.16862745098039217,\n",
       " 0.6039215686274509,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.054901960784313725,\n",
       " 0.00392156862745098,\n",
       " 0.6039215686274509,\n",
       " 0.9921568627450981,\n",
       " 0.35294117647058826,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.5450980392156862,\n",
       " 0.9921568627450981,\n",
       " 0.7450980392156863,\n",
       " 0.00784313725490196,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.043137254901960784,\n",
       " 0.7450980392156863,\n",
       " 0.9921568627450981,\n",
       " 0.27450980392156865,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.13725490196078433,\n",
       " 0.9450980392156862,\n",
       " 0.8823529411764706,\n",
       " 0.6274509803921569,\n",
       " 0.4235294117647059,\n",
       " 0.00392156862745098,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.3176470588235294,\n",
       " 0.9411764705882353,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.4666666666666667,\n",
       " 0.09803921568627451,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.17647058823529413,\n",
       " 0.7294117647058823,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.5882352941176471,\n",
       " 0.10588235294117647,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.06274509803921569,\n",
       " 0.36470588235294116,\n",
       " 0.9882352941176471,\n",
       " 0.9921568627450981,\n",
       " 0.7333333333333333,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.9764705882352941,\n",
       " 0.9921568627450981,\n",
       " 0.9764705882352941,\n",
       " 0.25098039215686274,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.1803921568627451,\n",
       " 0.5098039215686274,\n",
       " 0.7176470588235294,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.8117647058823529,\n",
       " 0.00784313725490196,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.15294117647058825,\n",
       " 0.5803921568627451,\n",
       " 0.8980392156862745,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9803921568627451,\n",
       " 0.7137254901960784,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.09411764705882353,\n",
       " 0.4470588235294118,\n",
       " 0.8666666666666667,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.788235294117647,\n",
       " 0.3058823529411765,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.09019607843137255,\n",
       " 0.25882352941176473,\n",
       " 0.8352941176470589,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.7764705882352941,\n",
       " 0.3176470588235294,\n",
       " 0.00784313725490196,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.07058823529411765,\n",
       " 0.6705882352941176,\n",
       " 0.8588235294117647,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.7647058823529411,\n",
       " 0.3137254901960784,\n",
       " 0.03529411764705882,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.21568627450980393,\n",
       " 0.6745098039215687,\n",
       " 0.8862745098039215,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9568627450980393,\n",
       " 0.5215686274509804,\n",
       " 0.043137254901960784,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.5333333333333333,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.9921568627450981,\n",
       " 0.8313725490196079,\n",
       " 0.5294117647058824,\n",
       " 0.5176470588235295,\n",
       " 0.06274509803921569,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0,\n",
       " 0.0]"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "flatten2d(x_train[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}