From 6565dd2dd80a88fef39ec16cb55bc9253d59ffda Mon Sep 17 00:00:00 2001
From: yashasvi <98yashasvi@gmail.com>
Date: Fri, 3 Apr 2020 01:31:04 +0530
Subject: [PATCH 1/8] Sqashing 6 commits

---
 Aggressive_Cows/Aggressive_Cows.java          |  76 ++++
 Boyer_Moore_Algorithm/Boyer_Moore.cs          | 110 +++++
 Johnson_Algorithm/Johnson_Algorithm.dart      | 180 ++++++++
 .../Logistic_Regression.ipynb                 | 418 ++++++++++++++++++
 .../Tree_Preorder_Traversal.kt                |  50 +++
 kosaraju_algorithm/kosaraju_algorithm.cpp     | 105 +++++
 6 files changed, 939 insertions(+)
 create mode 100644 Aggressive_Cows/Aggressive_Cows.java
 create mode 100644 Boyer_Moore_Algorithm/Boyer_Moore.cs
 create mode 100644 Johnson_Algorithm/Johnson_Algorithm.dart
 create mode 100644 Machine_Learning/Logistic_Regression/Logistic_Regression.ipynb
 create mode 100644 Tree_Preorder_Traversal/Tree_Preorder_Traversal.kt
 create mode 100644 kosaraju_algorithm/kosaraju_algorithm.cpp

diff --git a/Aggressive_Cows/Aggressive_Cows.java b/Aggressive_Cows/Aggressive_Cows.java
new file mode 100644
index 0000000000..1cdb49723f
--- /dev/null
+++ b/Aggressive_Cows/Aggressive_Cows.java
@@ -0,0 +1,76 @@
+import java.util.Arrays;
+import java.util.Scanner;
+
+public class Aggressive_Cows {
+
+    public static void main(String[] args) {
+
+        /* in this problem we have to maximise the minimum distance between the
+         * cows since minimum distance is being linearly checked we apply
+         * binary search on minimum distance */
+
+        Scanner scn = new Scanner(System.in);
+        int nos = scn.nextInt(); // input for no. of stalls
+        int noc = scn.nextInt(); // input for no. of cows
+
+        int[] arr = new int[nos]; // storing the position of stalls in an array
+
+        for (int i = 0; i < arr.length; i++) {
+            arr[i] = scn.nextInt();
+        }
+
+        Arrays.sort(arr); // to sort the positions of stalls in ascending order
+
+        int finalAns = 0;
+
+        int lo = 0;
+        int hi = arr[arr.length - 1] - arr[0];
+
+        while (lo <= hi) {
+
+            int mid = (lo + hi) / 2;
+
+            if (isItPossible(nos, noc, arr, mid)) {
+
+                finalAns = mid;
+                lo = mid + 1;
+
+            }else{
+                hi = mid - 1;
+            }
+        }
+
+        System.out.println(finalAns);
+
+    }
+
+    // function to check if a particular arrangement of cows is possible or not
+    private static boolean isItPossible(int nos, int noc, int[] arr, int mid) {
+
+        int cowsPlaced = 1;
+        int lastCowPos = arr[0]; // position at which cow is placed
+
+        for (int i = 1; i < arr.length; i++) {
+
+            if (arr[i] - lastCowPos >= mid) {
+                cowsPlaced++;
+                lastCowPos = arr[i];
+
+                if (cowsPlaced == noc) {
+                    return true;
+                }
+            }
+        }
+        return false;
+    }
+}
+
+// SAMPLE INPUT: 5 3
+// 1
+// 2
+// 8
+// 4
+// 9
+//
+// Output:
+// 3
diff --git a/Boyer_Moore_Algorithm/Boyer_Moore.cs b/Boyer_Moore_Algorithm/Boyer_Moore.cs
new file mode 100644
index 0000000000..391afe81aa
--- /dev/null
+++ b/Boyer_Moore_Algorithm/Boyer_Moore.cs
@@ -0,0 +1,110 @@
+// C# Program for Boyer Moore String Matching Algorithm 
+
+using System;
+
+public class Algorithm
+{ 
+    static int CHARACTERS = 256; 
+    
+    // Getting maximum of two integers 
+    static int max (int a, int b) { return (a > b)? a: b; } 
+    
+    // Bad Character Pre-Processing Function
+    static void badChar( char []str, int size,int []badCharacter) 
+    { 
+        int i; 
+
+        // Initializing all occurences to -1 
+        for (i = 0; i < CHARACTERS; i++)
+            badCharacter[i] = -1; 
+
+        // Filling the Actual Value
+        for (i = 0; i < size; i++) 
+            badCharacter[(int) str[i]] = i;
+    } 
+
+    // Pattern Searching Function
+    static void search( char []txt, char []pat) 
+    {
+        int m = pat.Length; 
+        int n = txt.Length;
+
+        int []Character = new int[CHARACTERS];
+        badChar(pat, m, Character); 
+
+        /* 
+            s is used to keep track of 
+            pattern shifting with respect to text
+        */
+
+        int s = 0; 
+
+        while(s <= (n - m))
+        {
+            int j = m - 1; 
+
+            while(j >= 0 && pat[j] == txt[s+j])
+                j--; 
+
+            /* 
+                If the pattern is present at current 
+                shift, then index j will become -1 after 
+                the above loop 
+            */
+
+            if (j < 0) 
+            { 
+                Console.WriteLine("Pattern occurs at index: " + s); 
+                s += (s+m < n)? m-Character[txt[s+m]] : 1; 
+            } 
+
+            else 
+                s += max(1, j - Character[txt[s+j]]); 
+
+        } 
+    } 
+
+    public static void Main() 
+    { 
+        Console.WriteLine("Enter The String Value: ");
+        String valueEntered = Console.ReadLine(); 
+        Console.WriteLine("Enter The Pattern To Search: ");
+        String pattern = Console.ReadLine();
+        Console.WriteLine(); 
+
+        char []txt = valueEntered.ToCharArray(); 
+        char []pat = pattern.ToCharArray(); 
+        search(txt, pat); 
+    } 
+} 
+
+/**
+
+Enter The String Value: 
+ABAAABCD
+Enter The Pattern To Search: 
+ABC
+
+Pattern occurs at index: 4
+
+--------------------------------------------------
+
+Enter The String Value: 
+AABAACAADAABAABA
+Enter The Pattern To Search: 
+AABA
+
+Pattern occurs at index: 0
+Pattern occurs at index: 9
+Pattern occurs at index: 12
+
+--------------------------------------------------
+
+Enter The String Value: 
+THIS IS A TEST TEXT
+Enter The Pattern To Search: 
+TEST
+
+Pattern occurs at index: 10
+
+*/
diff --git a/Johnson_Algorithm/Johnson_Algorithm.dart b/Johnson_Algorithm/Johnson_Algorithm.dart
new file mode 100644
index 0000000000..e1cbc70f97
--- /dev/null
+++ b/Johnson_Algorithm/Johnson_Algorithm.dart
@@ -0,0 +1,180 @@
+/*
+Johnson’s algorithm for All-pairs shortest paths
+
+Given a weighted Directed Graph where the weights may be negative,
+find the shortest path between every pair of vertices in the Graph using
+Johnson’s Algorithm.
+ */
+
+int MAX_INT = 9223372036854775807;
+
+int minDistance(distance, visited)
+{
+    var minimum  = MAX_INT;
+    var minVertex = 0;
+
+    var v = distance.length;
+
+    for(var vertex = 0; vertex < v; vertex++)
+    {
+        if( (minimum > distance[vertex]) && (visited[vertex] == false) )
+        {
+            minimum = distance[vertex];
+            minVertex = vertex;
+        }
+    }
+    return minVertex;
+}
+
+void Dijkstra(graph, modified, src)
+{
+    var num_vertices = graph.length;
+    var distance = new List(num_vertices);
+
+    var visited = new List(num_vertices);
+
+    for (var i = 0; i < num_vertices; i++)
+    {
+        distance[i] = MAX_INT;
+        visited[i] = false;
+    }
+
+    distance[src] = 0;
+
+    for(var count = 0; count < num_vertices; count++)
+    {
+        var curVertex = minDistance(distance, visited);
+        visited[curVertex] = true;
+        for(var vertex = 0; vertex < num_vertices; vertex++)
+        {
+            if ((visited[vertex] == false) && (distance[vertex] > (distance[curVertex] +
+            modified[curVertex][vertex])) && (graph[curVertex][vertex] != 0))
+            {
+               distance[vertex] = (distance[curVertex] + modified[curVertex][vertex]);
+            }
+        }
+    }
+
+    for(var vertex = 0; vertex < num_vertices; vertex++)
+    {
+        print('Vertex   ${vertex} :  ${distance[vertex]}');
+    }
+}
+
+List BellmanFord(edges, graph, num_vertices)
+{
+    var distance = new List(num_vertices+1);
+
+    for(var i = 0; i <= num_vertices; i++)
+    {
+        distance[i]=MAX_INT;
+    }
+
+    distance[num_vertices] = 0;
+
+    for(var i = 0; i < num_vertices; i++)
+    {
+        edges.add([num_vertices, i, 0]);
+    }
+
+    for(var i = 0; i < num_vertices; i++)
+    {
+        for(var j in edges)
+        {
+
+            if((distance[j[0]] != MAX_INT) && (distance[j[0]] + j[2] < distance[j[1]]) )
+            {
+                distance[j[1]] = distance[j[0]] + j[2];
+            }
+        }
+    }
+
+    return distance;
+}
+
+
+void JohnsonAlgorithm(graph)
+{
+    var edges = new List();
+
+    for(var i = 0; i < graph.length ; i++)
+    {
+        for(var j = 0; j < graph[i].length; j++)
+        {
+          if(graph[i][j] != 0)
+          {
+              edges.add([i, j, graph[i][j]]);
+          }
+        }
+    }
+
+    var modifiedwei = BellmanFord(edges, graph, graph.length);
+
+    var modified = [[0,0,0,0],
+                        [0,0,0,0],
+                        [0,0,0,0],
+                        [0,0,0,0]];
+
+    for(var i = 0; i < graph.length; i++)
+    {
+        for(var j = 0; j < graph[i].length; j++)
+        {
+            if(graph[i][j] != 0)
+            {
+                modified[i][j] = (graph[i][j] +
+                modifiedwei[i] - modifiedwei[j]);
+            }
+        }
+    }
+
+    print ('Modified Graph: ${modified}');
+
+    for(var src = 0; src < graph.length; src++)
+    {
+        print ('\nShortest Distance with vertex ${src} as the source:\n');
+        Dijkstra(graph, modified, src);
+    }
+}
+
+void main() {
+
+    var graph = [[0, -8, 2, 4],
+                  [0, 0, 2, 6],
+                  [0, 0, 0, 2],
+                  [0, 0, 0, 0]];
+
+    JohnsonAlgorithm(graph);
+}
+
+/*
+Modified Graph: [[0, 0, 8, 8], [0, 0, 0, 2], [0, 0, 0, 0], [0, 0, 0, 0]]
+
+Shortest Distance with vertex 0 as the source:
+
+Vertex   0 :  0
+Vertex   1 :  0
+Vertex   2 :  0
+Vertex   3 :  0
+
+Shortest Distance with vertex 1 as the source:
+
+Vertex   0 :  9223372036854775807
+Vertex   1 :  0
+Vertex   2 :  0
+Vertex   3 :  0
+
+Shortest Distance with vertex 2 as the source:
+
+Vertex   0 :  9223372036854775807
+Vertex   1 :  9223372036854775807
+Vertex   2 :  0
+Vertex   3 :  0
+
+Shortest Distance with vertex 3 as the source:
+
+Vertex   0 :  9223372036854775807
+Vertex   1 :  9223372036854775807
+Vertex   2 :  -9223372036854775801
+Vertex   3 :  0
+
+ */
diff --git a/Machine_Learning/Logistic_Regression/Logistic_Regression.ipynb b/Machine_Learning/Logistic_Regression/Logistic_Regression.ipynb
new file mode 100644
index 0000000000..669cfdb7fc
--- /dev/null
+++ b/Machine_Learning/Logistic_Regression/Logistic_Regression.ipynb
@@ -0,0 +1,418 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import matplotlib.pyplot as plt\n",
+    "import pandas as pd\n",
+    "%matplotlib inline"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "plt.style.use('seaborn')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "(500, 2)\n",
+      "(500, 2)\n"
+     ]
+    }
+   ],
+   "source": [
+    "mean_01 = np.array([1,0.5])\n",
+    "cov_01 = np.array([[1,0.1],[0.1,1.2]])\n",
+    "\n",
+    "mean_02 = np.array([4,5])\n",
+    "cov_02 = np.array([[1.21,0.1],[0.1,1.3]])\n",
+    "\n",
+    "\n",
+    "# Normal Distribution\n",
+    "dist_01 = np.random.multivariate_normal(mean_01,cov_01,500)\n",
+    "dist_02 = np.random.multivariate_normal(mean_02,cov_02,500)\n",
+    "\n",
+    "print(dist_01.shape)\n",
+    "print(dist_02.shape)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 576x396 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.figure(0)\n",
+    "\n",
+    "plt.scatter(dist_01[:,0],dist_01[:,1],label='Class 0')\n",
+    "plt.scatter(dist_02[:,0],dist_02[:,1],color='r',marker='^',label='Class 1')\n",
+    "plt.xlim(-5,10)\n",
+    "plt.ylim(-5,10)\n",
+    "plt.xlabel('x1')\n",
+    "plt.ylabel('x2')\n",
+    "plt.legend()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "(500, 1)\n",
+      "(500, 1)\n",
+      "(1000, 3)\n"
+     ]
+    }
+   ],
+   "source": [
+    "y1 = np.zeros((500,1))\n",
+    "print(y1.shape)\n",
+    "\n",
+    "y2 = np.ones((500,1))\n",
+    "print(y2.shape)\n",
+    "\n",
+    "data1 = np.hstack((dist_01,y1))\n",
+    "data2 = np.hstack((dist_02,y2))\n",
+    "data = np.vstack((data1,data2))\n",
+    "\n",
+    "print(data.shape)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "[[ 4.61546313  2.66193635  1.        ]\n",
+      " [ 4.14312954  4.21854945  1.        ]\n",
+      " [ 3.5360558   5.44785463  1.        ]\n",
+      " [-0.30283189  1.49333376  0.        ]\n",
+      " [ 3.78745303  6.42600424  1.        ]\n",
+      " [ 1.88347777  4.75120101  1.        ]\n",
+      " [ 4.62331128  4.57041199  1.        ]\n",
+      " [ 3.25889833  6.95708425  1.        ]\n",
+      " [ 0.8891156  -1.34359916  0.        ]\n",
+      " [ 3.17329477  3.51864788  1.        ]]\n"
+     ]
+    }
+   ],
+   "source": [
+    "np.random.shuffle(data)\n",
+    "print(data[:10])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "split = int(0.8*data.shape[0])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "(800, 2) (200, 2)\n",
+      "(800,) (200,)\n"
+     ]
+    }
+   ],
+   "source": [
+    "X_train = data[:split,:-1]\n",
+    "X_test = data[split:,:-1]\n",
+    "\n",
+    "Y_train = data[:split,-1]\n",
+    "Y_test  = data[split:,-1]\n",
+    "\n",
+    "print(X_train.shape,X_test.shape)\n",
+    "print(Y_train.shape,Y_test.shape)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def hypothesis(x,w,b):\n",
+    "    '''accepts input vector x, input weight vector w and bias b'''\n",
+    "    \n",
+    "    h = np.dot(x,w) + b\n",
+    "    return sigmoid(h)\n",
+    "\n",
+    "def sigmoid(z):\n",
+    "    return 1.0/(1.0 + np.exp(-1.0*z))\n",
+    "\n",
+    "def error(y_true,x,w,b):\n",
+    "    \n",
+    "    m = x.shape[0]\n",
+    "    \n",
+    "    err = 0.0\n",
+    "    \n",
+    "    for i in range(m):\n",
+    "        hx = hypothesis(x[i],w,b) \n",
+    "        err += y_true[i]*np.log2(hx) + (1-y_true[i])*np.log2(1-hx)\n",
+    "        \n",
+    "    \n",
+    "    return -err/m\n",
+    "\n",
+    "\n",
+    "def get_grads(y_true,x,w,b):\n",
+    "    \n",
+    "    grad_w = np.zeros(w.shape)\n",
+    "    grad_b = 0.0\n",
+    "    \n",
+    "    m = x.shape[0]\n",
+    "    \n",
+    "    for i in range(m):\n",
+    "        hx = hypothesis(x[i],w,b)\n",
+    "        \n",
+    "        grad_w += -1*(y_true[i] - hx)*x[i]\n",
+    "        grad_b +=  -1*(y_true[i]-hx)\n",
+    "        \n",
+    "    \n",
+    "    grad_w /= m\n",
+    "    grad_b /= m\n",
+    "    \n",
+    "    return [grad_w,grad_b] #returning python list\n",
+    "\n",
+    "\n",
+    "# One Iteration of Gradient Descent\n",
+    "def grad_descent(x,y_true,w,b,learning_rate=0.1):\n",
+    "    \n",
+    "    err = error(y_true,x,w,b)\n",
+    "    [grad_w,grad_b] = get_grads(y_true,x,w,b)\n",
+    "    \n",
+    "    w = w - learning_rate*grad_w\n",
+    "    b = b - learning_rate*grad_b\n",
+    "    \n",
+    "    return err,w,b\n",
+    "    \n",
+    "def predict(x,w,b):\n",
+    "    \n",
+    "    confidence = hypothesis(x,w,b)\n",
+    "    if confidence<0.5:\n",
+    "        return 0\n",
+    "    else:\n",
+    "        return 1\n",
+    "    \n",
+    "def accuracy(x_tst,y_tst,w,b):\n",
+    "    \n",
+    "    y_pred = []\n",
+    "    \n",
+    "    for i in range(y_tst.shape[0]):\n",
+    "        p = predict(x_tst[i],w,b)\n",
+    "        y_pred.append(p)\n",
+    "        \n",
+    "    y_pred = np.array(y_pred)\n",
+    "    \n",
+    "    return  float((y_pred==y_tst).sum())/y_tst.shape[0]\n",
+    "        \n",
+    "    \n",
+    "    "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "loss = []\n",
+    "acc = []\n",
+    "\n",
+    "W = 2*np.random.random((X_train.shape[1],))\n",
+    "b = 5*np.random.random()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "for i in range(300):\n",
+    "    l,W,b = grad_descent(X_train,Y_train,W,b,learning_rate=0.5)\n",
+    "    acc.append(accuracy(X_test,Y_test,W,b))\n",
+    "    loss.append(l)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 576x396 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.plot(loss)\n",
+    "plt.ylabel(\"negative of log likelihood\")\n",
+    "plt.xlabel(\"epoch\")\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe8AAAFYCAYAAAB6RnQAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAgAElEQVR4nO3df3RU9Z3/8ddkJr9/T5gJQsAEhAYCuEbEpVRUCmyL2x/a1Wa71B/4q8cf6+lKtxo9S9UCWrHaY/fsiivneBAwR0tP9Wst1lZriylY0QhBRdBAAkhm8mPIJJOfc79/BKaMJGGAuXNnMs/HOZ6TO7/yzptrXvncz733YzMMwxAAAEgYKVYXAAAATg/hDQBAgiG8AQBIMIQ3AAAJhvAGACDBEN4AACQYh5kfvmfPHt122226/vrrtXTp0rDn3n77bf385z+X3W7X/Pnzdfvtt4/4WR5PR1RrKyzMUltbV1Q/M5HRj3D0Ixz9CEc/wtGPcNHsh8uVO+Tjpo28u7q69NBDD2nu3LlDPv/Tn/5UTz75pDZt2qStW7dq7969ZpUyJIfDHtPvF+/oRzj6EY5+hKMf4ehHuFj0w7TwTktL09NPPy23233Sc42NjcrPz9c555yjlJQUXXrppaqtrTWrFAAARhXTDps7HA45HEN/vMfjkdPpDG07nU41NjaO+HmFhVlR/2tmuMMRyYp+hKMf4ehHOPoRjn6EM7sfps55R1O051Ncrtyoz6MnMvoRjn6Eox/h6Ec4+hEumv2I+Zz3SNxut7xeb2j7yJEjQx5eBwAAJ7MkvEtKSuT3+9XU1KT+/n698cYbmjdvnhWlAACQcEw7bL5r1y498sgjOnjwoBwOh7Zs2aIFCxaopKREixYt0k9+8hPdfffdkqQlS5aorKzMrFIAABhVTAvvGTNmaP369cM+f9FFF6mmpsasbw8AwKjFHdYAAEgwhDcAAAmG8AYAIMEkzHXeABLfZ4eP6sP9bVaXcdays9PV2dljdRlxg34MKshJ09yKsTH5XoQ3gJgI9PTriRfq1NHVZ3UpgGlmTCpSLO5aQngDiInX3mlUR1efLq8cr/Mnj7G6nLOSn58pny9gdRlxg34MKshJU15WWky+F+ENHNPXH9Rjz7+ng95OpaTYFAwaVpcUN6LRj0DPgPKyUnX1ZZOVkZbYv3q4HWg4+hF7if1/EBBFb75/UHuafCrMTVdedpr6B4JWlxQ3HPaUs+5HYa5N35hXmvDBDcQD/i9CUurpG9COjz2hQDIk/b+3G5SRZtdPbrhIk84tYiRxAkZWQHwhvJGUXnxzn/7wbtNJj3/7kjLlxmjOCgDOFOGNpNPcHtCb7x2UqyBD35z393vqp6XadcGUxD6RCkByILwRtz7Y16IX3tirgSifONbV3aeBoKEr50/SP06PzTWZABBNhDfiUv9AUM+99rFajnZH/zC2zaZ/OG+M5kwrju7nAkCMEN6IO3sa2/X+J155fd1aeGGJvrdoqtUlAUBcIbwRV97b49GTm3dKktLT7PrneaXWFgQAcYjwRtwYCAb14p/2KcVm05Xzy/SliYUxu1sRACQSwhsx81bdIf1l5+Fhn+/tHdDhli7NP3+crphbGrvCACDBEN6IibaOHm38/R719geVYrMN+7rC3HR96ytlwz4PACC8ESMvb/1Mvf1BXf/1cs0/f5zV5QBAQkuxugCMfp+3dumtusM6pyhL82ZyXTUAnC3CG6bb/NanChqGrpo/SfYUdjkAOFscNkfU9fUPaNMf9srn75FhSO/v9arsnDxVTnVZXRoAjAqEN6LuD+8e1JvvHQxt21Ns+u6C82Qb4UQ1AEDkCG+MqH8gqBZfd8Sv7xsI6pXaBmWmO/TAsouUme6Qw56i9FS7eUUCQJIhvDGiJ3+1Uzs/bTnt933n0kkak59pQkUAAMIbw6pvaNXOT1s0fky2Jo/Pi/h92RmpWjR7gomVAUByI7wRZscejz7YNzjS/uhAmyTppn+ernPH5lpZFgDgBIQ3Qnz+Hq19uV69fcHQY3MrxhLcABBnCO8k0ts3oP6B4JDP+QN9+s3WBvX2BXX1ZZN1wVSXbJLGFGTEtkgAwCkR3kni4wNtWvP8+xoIGiO+rrgwU4sumiCHnZupAEC8IryTgGEYqvnjXg0EDf3DeWM01OXWaWkO9fcNENwAkAAI71Hur7s/1+7P2tTweYfmTHPrB9+aMeTrXK5ceTwdMa4OAHAmCO9RrN3fo7Uv7ZYkOewpunL+JIsrAgBEA+E9iu1uaJUkLb5ogr56YYlcBdw0BQBGAyY3R7HdDYPXaX95xliCGwBGEcJ7lDIMQ/UNrcrLSlWJO8fqcgAAUUR4j1KHWrrk8/dqWqlTKazmBQCjCuE9Sr2987AkqaLUaXElAIBoI7xHodaj3Xr93SYV5qbr4uluq8sBAEQZZ5uPIoGefv285n0daulUX39Q3/5KmVIdrKMNAKMNI+9R5LV3GrXv0FFlpDlUOdWlL88ca3VJAAATMPIeBY60demTRp9+t/2A8rJStfLmi5WRxj8tAIxWpv6GX7Vqlerq6mSz2VRdXa1Zs2aFnnv99df1P//zP0pLS9MVV1yhpUuXmlnKqNXbN6CfbXxPbR09kqR/uXQywQ0Ao5xpv+W3b9+u/fv3q6amRvv27VN1dbVqamokScFgUA899JB+/etfq6CgQDfffLMWLlyosWM5zHu6/rCjSW0dPbp4erEumDJGs8s5QQ0ARjvT5rxra2u1cOFCSdLkyZPl8/nk9/slSW1tbcrLy5PT6VRKSor+8R//UW+//bZZpYxagZ5+/bZ2v7LSHVq6eKrmTCvmmm4ASAKmjby9Xq8qKipC206nUx6PRzk5OXI6ners7FRDQ4PGjx+vbdu2ac6cOSN+XmFhlhxRPnPa5cqN6ufFWt0ejzq7+3XlZeepdMLZX8+d6P2INvoRjn6Eox/h6Ec4s/sRs8lRwzBCX9tsNj388MOqrq5Wbm6uSkpKTvn+trauqNYzGpbA3PVJsySpOD/9rH+W0dCPaKIf4ehHOPoRjn6Ei2Y/hvsjwLTwdrvd8nq9oe3m5ma5XK7Q9pw5c7Rx40ZJ0mOPPabx48ebVcqo1eTplCSVuLh3OQAkE9PmvOfNm6ctW7ZIkurr6+V2u5WT8/eQuemmm9TS0qKuri698cYbmjt3rlmljFqNHr8c9hQVO1kxDACSiWkj78rKSlVUVKiqqko2m00rVqzQ5s2blZubq0WLFumaa67RsmXLZLPZdMstt8jp5B7cpyMYNHTI26lxY7JkT+FeOwCQTEyd816+fHnYdnl5eejrxYsXa/HixWZ++1HtSFuX+vqDHDIHgCTEkC1BHWS+GwCSFuGdoD4+0C5JmuAmvAEg2RDeCajF160/1R1UUV6Gpk4osLocAECMEd4J6Dd/+Uz9A4a+fUmZUh38EwJAsuE3f4I56O3U1l2HNd6VrbkV3AseAJIR4Z1gNv9pnwxD+s78yUpJ4T7mAJCMCO8EsrfJp/c+8eq8knydf16R1eUAACxCeCcIwzD04pt7JQ2u2W1j9TAASFqEd4LY+WmL9jT5dP7kIs4wB4AkR3gngKBh6MU3P5VN0ncunWx1OQAAixHeCWBb/RE1efyaO2OsSrgpCwAkPcI7Aby09TM57DZ9+ytlVpcCAIgDhHec83X26khbQBWlTo0pYOlPAADhHfeaPH5J0oRiDpcDAAYR3nHuYPNgeLN6GADgOMI7zjUeH3lzohoA4BjCO841eTrlsKfIXch8NwBgEOEdx4JBQ4e8nRo/Jlv2FP6pAACDSIQ4dqStS339QZW4s60uBQAQRwjvOPb+Xq8kaWJxrsWVAADiCeEdp7q6+/Tb2v3KSnfoyzNYtxsA8HeEd5x6ddsBdXb3a8ncc5WdkWp1OQCAOEJ4x6G2jh79/p1GFeSk6asXllhdDgAgzhDecejlrZ+ptz+ob36lTOmpdqvLAQDEGcI7znR19+nPHxxWcWGmLpl1jtXlAADiEOEdZz7c366BoKGLpxdzbTcAYEikQ5zZvb9VklRR5rS4EgBAvCK848zuhjZlpNlVdk6e1aUAAOIU4R1HWnzdOtLapfKJhXLY+acBAAyNhIgjHx1okyRNO7fQ4koAAPGM8I4jjcfW7i4bxyFzAMDwCO840nRs7e7xY1iIBAAwPMI7jjQ1+zUmP0OZ6Q6rSwEAxDHCO074Ont1tKtPJa4cq0sBAMQ5wjtONB2b7y5xE94AgJER3nHi+Hx3iYv5bgDAyJhctZhhGPrNXz7Ttg+bJUkTGHkDAE6B8LbY/iMdemlrgyTJmZcud2GmtQUBAOIe4W2x3Q2DN2b5/uKpmjfzHBYjAQCcEklhsfrPBhciufBLbqWxdjcAIAKEt4V6+wb0SZNPE9w5ystOs7ocAECCILwt9EmTT/0DQU0v5V7mAIDImTrnvWrVKtXV1clms6m6ulqzZs0KPbdhwwa99NJLSklJ0YwZM3TfffeZWUpc2t1wbO3uUtbuBgBEzrSR9/bt27V//37V1NRo5cqVWrlyZeg5v9+vZ555Rhs2bNCmTZu0b98+vf/++2aVErd2N7TJYbdpyoQCq0sBACQQ08K7trZWCxculCRNnjxZPp9Pfv/gjUhSU1OVmpqqrq4u9ff3KxAIKD8/36xS4lJHV68OHOnQeePzlc6JagCA02DaYXOv16uKiorQttPplMfjUU5OjtLT03X77bdr4cKFSk9P1xVXXKGysrIRP6+wMEsOR3RDzuXKjernnY6P3j8oQ9JFFedYWseJ4qWOeEE/wtGPcPQjHP0IZ3Y/Ynadt2EYoa/9fr+eeuop/e53v1NOTo6uu+46ffTRRyovLx/2/W1tXVGtx+XKlcfTEdXPPB1//eCQJKnUnW1pHcdZ3Y94Qz/C0Y9w9CMc/QgXzX4M90eAaYfN3W63vF5vaLu5uVkul0uStG/fPk2YMEFOp1NpaWmaPXu2du3aZVYpcaeru1/vftysnMxUnVvMX6sAgNNjWnjPmzdPW7ZskSTV19fL7XYrJ2fwvt3jx4/Xvn371N3dLUnatWuXSktLzSol7ry6bb86u/v1tYsnKiXFZnU5AIAEY9ph88rKSlVUVKiqqko2m00rVqzQ5s2blZubq0WLFunGG2/UtddeK7vdrgsuuECzZ882q5S40u7v0e/faVRBTpq+emGJ1eUAABKQqXPey5cvD9s+cU67qqpKVVVVZn77uPTS1gb19gdV9ZUyzjIHAJwR7rAWQ0dau/TW+4dU7MzSJbPOsbocAECCIrxj6K26Qwoahq68pIzVwwAAZ4wEiaEDzYM3qZlRVmRxJQCAREZ4x1BTs19FeenKymAZdQDAmSO8Y+RoV698nb0qceVYXQoAIMER3jFy8Ngh8xI34Q0AODuEd4w0eToliZE3AOCsEd4x0ug5NvJ2ZVtcCQAg0RHeMdLU7JfDblOxM8vqUgAACY7wjgHDMPR5a5eKnVly2Gk5AODskCQx4A/0qbt3QO6CTKtLAQCMAoR3DHjaB1dPcxHeAIAoILxjoLm9SxLhDQCIDsI7Bv4+8s6wuBIAwGhAeMeApz0giZE3ACA6CO8Y8LYHZJM0Jp+RNwDg7BHeMdDcHlBBbrpSHXarSwEAjAKEt8n6+oNqO9rDIXMAQNREFN6GYZhdx6jVcrRbhsQ13gCAqIkovC+//HI9/vjjamxsNLueUefzlmOXiRUS3gCA6IgovF944QW5XC5VV1frhhtu0Msvv6ze3l6zaxsVmo4tSDKB1cQAAFESUXi7XC4tXbpU69ev109+8hNt2rRJl1xyiR5//HH19PSYXWNCa2I1MQBAlEV8wto777yje++9VzfffLMqKyu1ceNG5eXl6a677jKzvoTX2OxXRppdRVwmBgCIEkckL1q0aJHGjx+va665Rg8++KBSU1MlSZMnT9brr79uaoGJrK9/QEdaA5o0Lk82m83qcgAAo0RE4f1///d/MgxDpaWlkqTdu3dr+vTpkqSNGzeaVlyiO+TtUtAwVOJmvhsAED0RHTbfvHmznnrqqdD22rVrtWbNGkliRDmCv5+sxnw3ACB6Igrvbdu2afXq1aHtJ554Qu+++65pRY0Wjc2D4T2eM80BAFEUUXj39fWFXRrW2dmp/v5+04oaLT5ubJc9xaZzi3OtLgUAMIpENOddVVWlJUuWaMaMGQoGg9q5c6fuuOMOs2tLaP5Anw583qGpEwqUnsY9zQEA0RNReF999dWaN2+edu7cKZvNpnvvvVc5ORwKHsmH+9tkSJpeWmh1KQCAUSbi67y7urrkdDpVWFioTz/9VNdcc42ZdSW83Q2tkqTpZU6LKwEAjDYRjbx/+tOfauvWrfJ6vZo4caIaGxu1bNkys2tLaLsbWpWZ7lDpWOa7AQDRFdHIe+fOnXr11VdVXl6uX/3qV1q3bp0CgYDZtSWs5vaAPO3dmnZuoewprLoKAIiuiJIlLS1N0uBZ54ZhaMaMGdqxY4ephSWy0CFz5rsBACaI6LB5WVmZNmzYoNmzZ+uGG25QWVmZOjo6zK4tYe1uaJMkTS9lvhsAEH0RhfcDDzwgn8+nvLw8vfLKK2ppadGtt95qdm0JKRg09GFDq4ry0lXMGt4AABNEFN6rVq3SfffdJ0n6xje+YWpBie5Ac4c6u/t1wVQXt44FAJgiojlvu92u2tpa9fT0KBgMhv7DyY4fMq/gkDkAwCQRjbxfeOEFPfvsszIMI/SYzWbThx9+aFphiar+s8GT1aady8lqAABzRBTeLEISmd6+AX3S5NMEd47ystOsLgcAMEpFFN6/+MUvhnz8rrvuimoxie6Tgz71DwQ5ZA4AMFXEc97H/wsGg9q2bRuXig1h92dc3w0AMF9EI+8vriA2MDCgO++885TvW7Vqlerq6mSz2VRdXa1Zs2ZJko4cOaLly5eHXtfY2Ki777474c9k339k8A+a80ryLa4EADCaRRTeX9Tf368DBw6M+Jrt27dr//79qqmp0b59+1RdXa2amhpJUnFxsdavXx/6rO9///tasGDBmZQSVzztAeVnpykj7YzaCgBARCJKmUsvvTTsmmWfz6crr7xyxPfU1tZq4cKFkqTJkyfL5/PJ7/eftJTor3/9a/3TP/2TsrOzT7f2uDIQDKrF16NJ4/KsLgUAMMpFFN4bN24MfW2z2ZSTk6O8vJFDyuv1qqKiIrTtdDrl8XhOCu8XXnhB69atO2UNhYVZcjjskZQbMZcreit+fd7SqaBhqGRsblQ/N5YStW6z0I9w9CMc/QhHP8KZ3Y+IwjsQCOg3v/mN7r77bknSvffeq2XLlmnKlCkRf6MTrxE/7r333tOkSZNOCvShtLV1Rfy9IuFy5crjid5Jdx8fW4wkL8MR1c+NlWj3I9HRj3D0Ixz9CEc/wkWzH8P9ERDR2eYPPPCALr300tD2d77zHT344IMjvsftdsvr9Ya2m5ub5XK5wl7z5ptvau7cuZGUEPea2weXSHUVcD9zAIC5IgrvgYEBzZ49O7Q9e/bsIUfSJ5o3b562bNkiSaqvr5fb7T5phL1z506Vl5efbs1xyUN4AwBiJKLD5rm5udq4caMuvvhiBYNB/fnPfz7lCWaVlZWqqKhQVVWVbDabVqxYoc2bNys3N1eLFi2SJHk8HhUVFZ39TxEHPO3dkghvAID5Igrv1atX67HHHtOmTZskDQbz6tWrT/m+E6/llnTSKPvll1+OtM6452kPKNWRovwcbosKADBXROHtdDp18803q7S0VJK0e/duOZ3cAvRE3vaAxuRnKIVlQAEAJotozvvxxx/XU089Fdpeu3at1qxZY1pRiSbQ06/O7n6NyeeQOQDAfBGF97Zt28IOkz/xxBOsNHaC1o4eSZIzL93iSgAAySCi8O7r61Nvb29ou7OzU/39/aYVlWjaj4V3YQ7hDQAwX0Rz3lVVVVqyZIlmzJihYDConTt36rrrrjO7toTR2jF4pnlhLuENADBfROF99dVXq7S0VG1tbbLZbFqwYIGeeuopXX/99SaXlxhCI2/CGwAQAxGF98qVK/WXv/xFXq9XEydOVGNjo5YtW2Z2bQmjzT84pVBAeAMAYiCiOe8PPvhAr776qsrLy/WrX/1K69atUyAQMLu2hNF2dPCwuZPwBgDEQEThnZY2eOORvr4+GYahGTNmaMeOHaYWlkja/D1KS01RZjrreAMAzBdR2pSVlWnDhg2aPXu2brjhBpWVlamjgxVkjmvv6FFhbkbYmucAAJglovB+4IEH5PP5lJeXp1deeUUtLS269dZbza4tIfT1B3W0q0/jxox8r3cAAKIlovC22WwqKCiQJH3jG98wtaBE4/NzpjkAILYimvPG8NpC4Z1hcSUAgGRBeJ+lNq7xBgDEGOF9llp8xy4T477mAIAYIbzPkqd98Hp3dwErigEAYoPwPkvHw3sM4Q0AiBHC+yw1tweUn52m9FS71aUAAJIE4X0WBoJBtfh65GLUDQCIIcL7LLQe7VHQMOQq4DIxAEDsEN5nofnYfDcjbwBALBHeZ8FDeAMALEB4nwXCGwBgBcL7LHjaB2/QQngDAGKJ8D5DhmGo4fBRZaU7lJ+TZnU5AIAkQnifIU97QF5ft6adW6gU1vEGAMQQ4X2Gdje0SZKmlxZaXAkAINkQ3mdod0OrJGl6mdPiSgAAyYbwPgPBoKEP97epKC+DBUkAADFHeJ8BT3tAnd39mjqhQDbmuwEAMUZ4n4Hj13cXOxl1AwBij/A+A9ycBQBgJcL7DHBzFgCAlQjvM3B8QRJOVgMAWIHwPgOe9oDSU+3KzUq1uhQAQBIivE+TYRjytAfkKsjgTHMAgCUI79PkD/Spu3eA+W4AgGUI79PEyWoAAKsR3qeJy8QAAFYjvE9Tk8cvSSouJLwBANYgvE/T7oY22VNsmjw+3+pSAABJymHmh69atUp1dXWy2Wyqrq7WrFmzQs8dPnxY//Ef/6G+vj5Nnz5dDz74oJmlREVnd58aPj+qKePzlZluausAABiWaSPv7du3a//+/aqpqdHKlSu1cuXKsOcffvhhLVu2TC+++KLsdrsOHTpkVilR89H+NhmGNL2UZUABANYxLbxra2u1cOFCSdLkyZPl8/nk9w/OFweDQb377rtasGCBJGnFihUaN26cWaVETX1DmyTCGwBgLdOO/Xq9XlVUVIS2nU6nPB6PcnJy1NraquzsbK1evVr19fWaPXu27r777hE/r7AwSw6HPao1uly5p/X6fYeOKjPdrjmzxsluH32nC5xuP0Y7+hGOfoSjH+HoRziz+xGziVvDMMK+PnLkiK699lqNHz9et9xyi958801ddtllw76/ra0rqvW4XLnyeDoifn1f/4AONvs1aXyeWls7o1pLPDjdfox29CMc/QhHP8LRj3DR7MdwfwSYNnx0u93yer2h7ebmZrlcLklSYWGhxo0bp4kTJ8put2vu3Ln65JNPzColKg55uxQ0DE1w5VhdCgAgyZkW3vPmzdOWLVskSfX19XK73crJGQw+h8OhCRMmqKGhIfR8WVmZWaVExfHru0tc2RZXAgBIdqYdNq+srFRFRYWqqqpks9m0YsUKbd68Wbm5uVq0aJGqq6t1zz33yDAMTZ06NXTyWrwKhbebkTcAwFqmznkvX748bLu8vDz09bnnnqtNmzaZ+e2jqql5MLzHjyG8AQDWGn2nTJuk0dOporwMZWVwcxYAgLUI7wj4Ont1tLNXEzhkDgCIA4R3BP7wbpMkaeqEAosrAQCA8D4ln79Hr71zQPnZabr8gvFWlwMAAOF9Km/VHVJvX1DfnFeq9LTo3uENAIAzQXifQnNbQJJUManI4koAABhEeJ9Ca0ePJKkwJ83iSgAAGER4n0K7v0c5malKjfKiKAAAnCnC+xTaOnpUmJtudRkAAIQQ3iMI9PSru3eA8AYAxBXCewRtx+a7C3IIbwBA/CC8R3A8vJ2MvAEAcYTwHkFo5E14AwDiCOE9gjb/scvECG8AQBwhvEdwfORNeAMA4gnhPYJ2whsAEIcI7xG0dfQozZGirHTW8AYAxA/CewTtnT0qyEmXzWazuhQAAEII72EYhiF/V59ys1KtLgUAgDCE9zC6ewc0EDSUnUl4AwDiC+E9DH+gT5KUS3gDAOIM4T2M4+HNyBsAEG8I72GERt7MeQMA4gzhPQx/FyNvAEB8IryH0cGcNwAgThHewzh+2DyH8AYAxBnCexiENwAgXhHewwiFd1aaxZUAABCO8B6Gv6tXkpSdwX3NAQDxhfAehj/Qr8x0hxx2WgQAiC8k0zD8gV7lZDLqBgDEH8J7CIZhyB/oU04m890AgPhDeA+hp29A/QMGZ5oDAOIS4T2E43dXI7wBAPGI8B6Cv5vwBgDEL8J7CKGRN4uSAADiEOE9BNbyBgDEM8J7CB3cGhUAEMcI7yF0Et4AgDhGeA+BkTcAIJ4R3kPghDUAQDwjvIfAcqAAgHhm6s27V61apbq6OtlsNlVXV2vWrFmh5xYsWKCxY8fKbrdLktasWaPi4mIzy4mYP9CnjDQ7i5IAAOKSaeG9fft27d+/XzU1Ndq3b5+qq6tVU1MT9pqnn35a2dnZZpVwxgbva86oGwAQn0wbWtbW1mrhwoWSpMmTJ8vn88nv95v17aKK8AYAxDPTRt5er1cVFRWhbafTKY/Ho5ycnNBjK1as0MGDB3XhhRfq7rvvls1mG/bzCguz5HDYo1qjy5V70mPdvf3q6w/KmZ855POjWbL9vKdCP8LRj3D0Ixz9CGd2P2K2YLVhGGHb//7v/65LLrlE+fn5uv3227VlyxZ97WtfG/b9bW1dUa3H5cqVx9Nx0uMtvm5JUprDNuTzo9Vw/UhW9CMc/QhHP8LRj3DR7MdwfwSYdtjc7XbL6/WGtpubm+VyuULb3/72t1VUVCSHw6H58+drz549ZpVyWjjTHAAQ70wL73nz5mnLli2SpPr6ernd7tAh846ODt14443q7e2VJL3zzjuaMmWKWaWcFsIbABDvTDtsXllZqYqKClVVVclms2nFihXavHmzcnNztWjRIs2fP1/f/e53lZ6erunTp494yDyWOgKDf1CwKAkAIF6ZOue9fPnysO3y8vLQ19ddd52uu+46M7/9GfPZYVEAAAnoSURBVOkM9EuSsglvAECc4i4kX9DRxcgbABDfCO8vOD7nzcgbABCvCO8vONI6eEnamPxMiysBAGBohPcXNHk6VZSXoayMmF0CDwDAaSG8T3C0q1e+zl6VuOLvfusAABxHeJ/gYPPgvddL3DmneCUAANYhvE/Q6OmUJJW4CG8AQPwivE/Q5GHkDQCIf4T3CZqa/XLYbSou5ExzAED8IryPCQYNHfJ2alxRthx22gIAiF+k1DHN7QH19gc1nvluAECcI7yPaTp2pvkE5rsBAHGO8D7m7yercY03ACC+Ed7HNB6/xpvD5gCAOEd4H3PQ06mczFTlZ6dZXQoAACMivCV19/aruT2gEle2bDab1eUAADAiwluDo26Jm7MAABID4S3pjfcOSpKmlhRYXAkAAKeW1OtetnX0qOHzo6rd9blKXDmq/JLL6pIAADilpA3v1qPduu/pv6q7d0CS9C+XTVIK890AgASQtOH9/Gsfq7t3QBeVu/WliQWaOanI6pIAAIhIUoZ3c1uXtmzbr7HOLN3yzemypzD1DwBIHEmZWkc7+5Rik7674DyCGwCQcJJy5H1eSb5eXP3Pam3ttLoUAABOW9IOO+0s+wkASFAkGAAACYbwBgAgwRDeAAAkGMIbAIAEQ3gDAJBgCG8AABIM4Q0AQIIhvAEASDCENwAACYbwBgAgwRDeAAAkGJthGIbVRQAAgMgx8gYAIMEQ3gAAJBjCGwCABEN4AwCQYAhvAAASDOENAECCIbwBAEgwDqsLsMKqVatUV1cnm82m6upqzZo1y+qSYmrbtm266667NGXKFEnS1KlTddNNN+k///M/NTAwIJfLpUcffVRpaWkWV2q+PXv26LbbbtP111+vpUuX6vDhw0P24aWXXtKzzz6rlJQUXXPNNbr66qutLj3qvtiLe+65R/X19SooKJAk3XjjjbrsssuSoheS9LOf/Uzvvvuu+vv7deutt2rmzJlJu29IJ/fjj3/8Y9LuH4FAQPfcc49aWlrU09Oj2267TeXl5bHdP4wks23bNuOWW24xDMMw9u7da1xzzTUWVxR7f/3rX40777wz7LF77rnH+O1vf2sYhmE89thjxoYNG6woLaY6OzuNpUuXGvfff7+xfv16wzCG7kNnZ6exePFi4+jRo0YgEDCuuOIKo62tzcrSo26oXvz4xz82/vjHP570utHeC8MwjNraWuOmm24yDMMwWltbjUsvvTRp9w3DGLofybx/vPLKK8batWsNwzCMpqYmY/HixTHfP5LusHltba0WLlwoSZo8ebJ8Pp/8fr/FVVlv27Zt+upXvypJuvzyy1VbW2txReZLS0vT008/LbfbHXpsqD7U1dVp5syZys3NVUZGhiorK7Vjxw6ryjbFUL0YSjL0QpIuuugi/eIXv5Ak5eXlKRAIJO2+IQ3dj4GBgZNelyz9WLJkiW6++WZJ0uHDh1VcXBzz/SPpwtvr9aqwsDC07XQ65fF4LKzIGnv37tUPfvAD/eu//qu2bt2qQCAQOkxeVFSUFD1xOBzKyMgIe2yoPni9XjmdztBrRuM+M1QvJOm5557Ttddeqx/+8IdqbW1Nil5Ikt1uV1ZWliTpxRdf1Pz585N235CG7ofdbk/a/eO4qqoqLV++XNXV1THfP5JyzvtERhLe2r20tFR33HGHvv71r6uxsVHXXntt2F/RydiToQzXh2Tpz7e+9S0VFBRo2rRpWrt2rX75y1/qggsuCHvNaO/F66+/rhdffFHr1q3T4sWLQ48n675xYj927dqV9PvH888/rw8//FA/+tGPwn7WWOwfSTfydrvd8nq9oe3m5ma5XC4LK4q94uJiLVmyRDabTRMnTtSYMWPk8/nU3d0tSTpy5MgpD5+OVllZWSf1Yah9Jhn6M3fuXE2bNk2StGDBAu3ZsyepevHnP/9Z//u//6unn35aubm5Sb9vfLEfybx/7Nq1S4cPH5YkTZs2TQMDA8rOzo7p/pF04T1v3jxt2bJFklRfXy+3262cnByLq4qtl156Sc8884wkyePxqKWlRVdddVWoL6+99pouueQSK0u0zJe//OWT+nD++edr586dOnr0qDo7O7Vjxw7Nnj3b4krNd+edd6qxsVHS4LkAU6ZMSZpedHR06Gc/+5meeuqp0NnUybxvDNWPZN4//va3v2ndunWSBqdiu7q6Yr5/JOWSoGvWrNHf/vY32Ww2rVixQuXl5VaXFFN+v1/Lly/X0aNH1dfXpzvuuEPTpk3Tj3/8Y/X09GjcuHFavXq1UlNTrS7VVLt27dIjjzyigwcPyuFwqLi4WGvWrNE999xzUh9+97vf6ZlnnpHNZtPSpUv1zW9+0+ryo2qoXixdulRr165VZmamsrKytHr1ahUVFY36XkhSTU2NnnzySZWVlYUee/jhh3X//fcn3b4hDd2Pq666Ss8991xS7h/d3d267777dPjwYXV3d+uOO+7QjBkzhvwdalY/kjK8AQBIZEl32BwAgERHeAMAkGAIbwAAEgzhDQBAgiG8AQBIMIQ3gLO2efNmLV++3OoygKRBeAMAkGCS/t7mQDJZv369Xn31VQ0MDGjSpEm66aabdOutt2r+/Pn66KOPJEmPP/64iouL9eabb+q///u/lZGRoczMTD300EMqLi5WXV2dVq1apdTUVOXn5+uRRx6R9Peb/+zbt0/jxo3TL3/5S9lsNit/XGDUYuQNJIkPPvhAv//977VhwwbV1NQoNzdXb7/9thobG3XVVVdp48aNmjNnjtatW6dAIKD7779fTz75pNavX6/58+friSeekCT96Ec/0kMPPaTnnntOF110kf70pz9JGlyp7qGHHtLmzZv1ySefqL6+3sofFxjVGHkDSWLbtm06cOCArr32WklSV1eXjhw5ooKCAs2YMUOSVFlZqWeffVYNDQ0qKirS2LFjJUlz5szR888/r9bWVh09elRTp06VJF1//fWSBue8Z86cqczMTEmDi990dHTE+CcEkgfhDSSJtLQ0LViwQP/1X/8VeqypqUlXXXVVaNswDNlstpMOd5/4+HB3VLbb7Se9B4A5OGwOJInKykq99dZb6uzslCRt2LBBHo9HPp9Pu3fvliTt2LFDX/rSl1RaWqqWlhYdOnRIklRbW6vzzz9fhYWFKigo0AcffCBJWrdunTZs2GDNDwQkMUbeQJKYOXOm/u3f/k3f//73lZ6eLrfbrYsvvljFxcXavHmzHn74YRmGoZ///OfKyMjQypUr9cMf/lBpaWnKysrSypUrJUmPPvqoVq1aJYfDodzcXD366KN67bXXLP7pgOTCqmJAEmtqatL3vvc9vfXWW1aXAuA0cNgcAIAEw8gbAIAEw8gbAIAEQ3gDAJBgCG8AABIM4Q0AQIIhvAEASDD/H0oQ082/kubpAAAAAElFTkSuQmCC\n",
+      "text/plain": [
+       "<Figure size 576x396 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "0.98\n"
+     ]
+    }
+   ],
+   "source": [
+    "plt.plot(acc)\n",
+    "plt.ylabel(\"accuracy\")\n",
+    "plt.xlabel(\"epoch\")\n",
+    "plt.show()\n",
+    "print(acc[-1])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 576x396 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.figure(0)\n",
+    "\n",
+    "plt.scatter(dist_01[:,0],dist_01[:,1],label='Class 0')\n",
+    "plt.scatter(dist_02[:,0],dist_02[:,1],color='r',marker='^',label='Class 1')\n",
+    "plt.xlim(-5,10)\n",
+    "plt.ylim(-5,10)\n",
+    "plt.xlabel('x1')\n",
+    "plt.ylabel('x2')\n",
+    "\n",
+    "x = np.linspace(-4,8,10)\n",
+    "y = -(W[0]*x + b)/W[1]\n",
+    "plt.plot(x,y,color='k')\n",
+    "\n",
+    "plt.legend()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "[0.76604691 1.49267387]\n",
+      "-5.424920453033558\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(W)\n",
+    "print(b)"
+   ]
+  },
+  {
+   "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": "3.6.8"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}
diff --git a/Tree_Preorder_Traversal/Tree_Preorder_Traversal.kt b/Tree_Preorder_Traversal/Tree_Preorder_Traversal.kt
new file mode 100644
index 0000000000..18bef0bd6c
--- /dev/null
+++ b/Tree_Preorder_Traversal/Tree_Preorder_Traversal.kt
@@ -0,0 +1,50 @@
+import java.util.*
+class Node<Int>(
+    var key:Int,
+    var left:Node<Int>? = null,
+    var right:Node<Int>? = null  
+)
+{
+    fun preorderTraversal()
+    {
+      print("$key ")
+      left?.preorderTraversal()
+      right?.preorderTraversal()
+    }
+}
+
+fun main() {
+    var read = Scanner(System.`in`)
+    println("Enter the size of Array:")
+    val arrSize = read.nextLine().toInt()
+    var arr = IntArray(arrSize)
+    val nodes = mutableListOf<Node<Int>>()
+    println("Enter the array respresentaion of binary tree")
+    for(i in 0 until arrSize)
+    {
+        arr[i] = read.nextLine().toInt()
+        nodes.add(Node(arr[i]))
+    }
+    for(i in 0..arrSize-2)
+    {
+        if((i*2)+1<arrSize && arr[(i*2)+1]!=-1)
+        nodes[i].left = nodes[(i*2)+1]
+        if((i*2)+2<arrSize && arr[(i*2)+2]!=-1)
+        nodes[i].right = nodes[(i*2)+2]
+    }
+    print("Pre Order traversal of tree is ")
+    nodes[0].preorderTraversal()
+}
+/*
+*Enter the size of Array:
+*7
+*Enter the array respresentaion of binary tree
+*1
+*2
+*3
+*4
+*5
+*-1
+*7
+*Pre Order traversal of tree is 1 2 4 5 3 7 
+*/
diff --git a/kosaraju_algorithm/kosaraju_algorithm.cpp b/kosaraju_algorithm/kosaraju_algorithm.cpp
new file mode 100644
index 0000000000..8b969d461e
--- /dev/null
+++ b/kosaraju_algorithm/kosaraju_algorithm.cpp
@@ -0,0 +1,105 @@
+//Strongly connected components(KOSARAJU ALGORITHM)
+#include <iostream>
+#include <vector>
+#include <stack>
+#include <unordered_set>
+using namespace std;
+
+//dfs to store vertices in a stack as per their finish time
+void dfs(vector<int>* edges, int start, unordered_set<int> &visited, stack<int> &finishStack) {
+    visited.insert(start);
+    for (int i = 0; i < edges[start].size(); i++) {
+        int adjacent = edges[start][i];
+        if (visited.count(adjacent) == 0) {
+            dfs(edges, adjacent, visited, finishStack);
+        }
+    }
+    finishStack.push(start);
+}
+
+//dfs of a transpose graph
+void dfs2(vector<int>* edges, int start, unordered_set<int>* component, unordered_set<int> & visited) {
+    visited.insert(start);
+    component->insert(start);
+    for (int i = 0; i < edges[start].size(); i++) {
+        int adjacent = edges[start][i];
+        if (visited.count(adjacent) == 0) {
+            dfs2(edges, adjacent, component, visited);
+        }
+    }
+}
+
+unordered_set<unordered_set<int>*>* getSCC(vector<int>* edges, vector<int>* edgesT, int n) {
+    unordered_set<int> visited;
+    stack<int> finishedVertices;
+    for (int i = 0; i < n; i++) {
+        if (visited.count(i) == 0) {
+            dfs(edges, i, visited, finishedVertices);
+        }
+    }
+    unordered_set<unordered_set<int>*>* output = new unordered_set<unordered_set<int>*>();
+    visited.clear();
+    while (finishedVertices.size() != 0) {
+        int element = finishedVertices.top();
+        finishedVertices.pop();
+        if (visited.count(element) != 0) {
+            continue;
+        }
+        unordered_set<int>* component = new unordered_set<int>();
+        dfs2(edgesT, element, component, visited);
+        output->insert(component);
+    }
+    return output;
+}
+
+int main() {
+    int n;    //no.of vertices
+    cout<<"Enter the no. of vertices : "<<endl;
+    cin >> n;
+    vector<int>* edges = new vector<int>[n];    //using a adjacency list to implement
+    vector<int>* edgesT = new vector<int>[n];    //transpose of a graph
+    int m;    // no.of edges
+    cout<<"Enter the no. of edges : "<<endl;
+    cin >> m;
+
+    for (int i = 0; i < m; i++) {
+        int j, k;
+        cin >> j >> k;
+        edges[j - 1].push_back(k - 1);
+        edgesT[k - 1].push_back(j - 1);
+    }
+
+    unordered_set<unordered_set<int>*>* components = getSCC(edges, edgesT, n);
+    unordered_set<unordered_set<int>*>::iterator it = components->begin();
+
+    while (it != components->end()) {
+        unordered_set<int>* component = *it;
+        unordered_set<int>::iterator it2 = component->begin();
+        while (it2 != component->end()) {
+            cout <<*it2  + 1<< " ";
+            it2++;
+        }
+        cout << endl;
+        delete component;
+        it++;
+    }
+
+    delete components;
+    delete [] edges;
+    delete [] edgesT;
+}
+
+/*INPUTS
+Enter the no. of vertices : 6
+Enter the no. of edges  :7
+1 2
+2 3
+3 4
+4 1
+3 5
+5 6
+6 5
+
+OUTPUT
+6 5
+2 3 4 1 */

From 6318684d1108a6bceabfc12c18bbab6c226f2c0f Mon Sep 17 00:00:00 2001
From: Yashasvi Sinha <59117774+heisastark@users.noreply.github.com>
Date: Fri, 3 Apr 2020 01:35:49 +0530
Subject: [PATCH 2/8] deleting accidentally added file

---
 Aggressive_Cows/Aggressive_Cows.java | 76 ----------------------------
 1 file changed, 76 deletions(-)
 delete mode 100644 Aggressive_Cows/Aggressive_Cows.java

diff --git a/Aggressive_Cows/Aggressive_Cows.java b/Aggressive_Cows/Aggressive_Cows.java
deleted file mode 100644
index 1cdb49723f..0000000000
--- a/Aggressive_Cows/Aggressive_Cows.java
+++ /dev/null
@@ -1,76 +0,0 @@
-import java.util.Arrays;
-import java.util.Scanner;
-
-public class Aggressive_Cows {
-
-    public static void main(String[] args) {
-
-        /* in this problem we have to maximise the minimum distance between the
-         * cows since minimum distance is being linearly checked we apply
-         * binary search on minimum distance */
-
-        Scanner scn = new Scanner(System.in);
-        int nos = scn.nextInt(); // input for no. of stalls
-        int noc = scn.nextInt(); // input for no. of cows
-
-        int[] arr = new int[nos]; // storing the position of stalls in an array
-
-        for (int i = 0; i < arr.length; i++) {
-            arr[i] = scn.nextInt();
-        }
-
-        Arrays.sort(arr); // to sort the positions of stalls in ascending order
-
-        int finalAns = 0;
-
-        int lo = 0;
-        int hi = arr[arr.length - 1] - arr[0];
-
-        while (lo <= hi) {
-
-            int mid = (lo + hi) / 2;
-
-            if (isItPossible(nos, noc, arr, mid)) {
-
-                finalAns = mid;
-                lo = mid + 1;
-
-            }else{
-                hi = mid - 1;
-            }
-        }
-
-        System.out.println(finalAns);
-
-    }
-
-    // function to check if a particular arrangement of cows is possible or not
-    private static boolean isItPossible(int nos, int noc, int[] arr, int mid) {
-
-        int cowsPlaced = 1;
-        int lastCowPos = arr[0]; // position at which cow is placed
-
-        for (int i = 1; i < arr.length; i++) {
-
-            if (arr[i] - lastCowPos >= mid) {
-                cowsPlaced++;
-                lastCowPos = arr[i];
-
-                if (cowsPlaced == noc) {
-                    return true;
-                }
-            }
-        }
-        return false;
-    }
-}
-
-// SAMPLE INPUT: 5 3
-// 1
-// 2
-// 8
-// 4
-// 9
-//
-// Output:
-// 3

From d9e94d3b7ae3f96eb47e3c02d1220deb37f92641 Mon Sep 17 00:00:00 2001
From: Yashasvi Sinha <59117774+heisastark@users.noreply.github.com>
Date: Fri, 3 Apr 2020 01:35:59 +0530
Subject: [PATCH 3/8] deleting accidentally added file

---
 Boyer_Moore_Algorithm/Boyer_Moore.cs | 110 ---------------------------
 1 file changed, 110 deletions(-)
 delete mode 100644 Boyer_Moore_Algorithm/Boyer_Moore.cs

diff --git a/Boyer_Moore_Algorithm/Boyer_Moore.cs b/Boyer_Moore_Algorithm/Boyer_Moore.cs
deleted file mode 100644
index 391afe81aa..0000000000
--- a/Boyer_Moore_Algorithm/Boyer_Moore.cs
+++ /dev/null
@@ -1,110 +0,0 @@
-// C# Program for Boyer Moore String Matching Algorithm 
-
-using System;
-
-public class Algorithm
-{ 
-    static int CHARACTERS = 256; 
-    
-    // Getting maximum of two integers 
-    static int max (int a, int b) { return (a > b)? a: b; } 
-    
-    // Bad Character Pre-Processing Function
-    static void badChar( char []str, int size,int []badCharacter) 
-    { 
-        int i; 
-
-        // Initializing all occurences to -1 
-        for (i = 0; i < CHARACTERS; i++)
-            badCharacter[i] = -1; 
-
-        // Filling the Actual Value
-        for (i = 0; i < size; i++) 
-            badCharacter[(int) str[i]] = i;
-    } 
-
-    // Pattern Searching Function
-    static void search( char []txt, char []pat) 
-    {
-        int m = pat.Length; 
-        int n = txt.Length;
-
-        int []Character = new int[CHARACTERS];
-        badChar(pat, m, Character); 
-
-        /* 
-            s is used to keep track of 
-            pattern shifting with respect to text
-        */
-
-        int s = 0; 
-
-        while(s <= (n - m))
-        {
-            int j = m - 1; 
-
-            while(j >= 0 && pat[j] == txt[s+j])
-                j--; 
-
-            /* 
-                If the pattern is present at current 
-                shift, then index j will become -1 after 
-                the above loop 
-            */
-
-            if (j < 0) 
-            { 
-                Console.WriteLine("Pattern occurs at index: " + s); 
-                s += (s+m < n)? m-Character[txt[s+m]] : 1; 
-            } 
-
-            else 
-                s += max(1, j - Character[txt[s+j]]); 
-
-        } 
-    } 
-
-    public static void Main() 
-    { 
-        Console.WriteLine("Enter The String Value: ");
-        String valueEntered = Console.ReadLine(); 
-        Console.WriteLine("Enter The Pattern To Search: ");
-        String pattern = Console.ReadLine();
-        Console.WriteLine(); 
-
-        char []txt = valueEntered.ToCharArray(); 
-        char []pat = pattern.ToCharArray(); 
-        search(txt, pat); 
-    } 
-} 
-
-/**
-
-Enter The String Value: 
-ABAAABCD
-Enter The Pattern To Search: 
-ABC
-
-Pattern occurs at index: 4
-
---------------------------------------------------
-
-Enter The String Value: 
-AABAACAADAABAABA
-Enter The Pattern To Search: 
-AABA
-
-Pattern occurs at index: 0
-Pattern occurs at index: 9
-Pattern occurs at index: 12
-
---------------------------------------------------
-
-Enter The String Value: 
-THIS IS A TEST TEXT
-Enter The Pattern To Search: 
-TEST
-
-Pattern occurs at index: 10
-
-*/

From 67e37d4aa58829bb6e571aa3595d02d75d5c2b4e Mon Sep 17 00:00:00 2001
From: Yashasvi Sinha <59117774+heisastark@users.noreply.github.com>
Date: Fri, 3 Apr 2020 01:36:07 +0530
Subject: [PATCH 4/8] deleting accidentally added file

---
 Johnson_Algorithm/Johnson_Algorithm.dart | 180 -----------------------
 1 file changed, 180 deletions(-)
 delete mode 100644 Johnson_Algorithm/Johnson_Algorithm.dart

diff --git a/Johnson_Algorithm/Johnson_Algorithm.dart b/Johnson_Algorithm/Johnson_Algorithm.dart
deleted file mode 100644
index e1cbc70f97..0000000000
--- a/Johnson_Algorithm/Johnson_Algorithm.dart
+++ /dev/null
@@ -1,180 +0,0 @@
-/*
-Johnson’s algorithm for All-pairs shortest paths
-
-Given a weighted Directed Graph where the weights may be negative,
-find the shortest path between every pair of vertices in the Graph using
-Johnson’s Algorithm.
- */
-
-int MAX_INT = 9223372036854775807;
-
-int minDistance(distance, visited)
-{
-    var minimum  = MAX_INT;
-    var minVertex = 0;
-
-    var v = distance.length;
-
-    for(var vertex = 0; vertex < v; vertex++)
-    {
-        if( (minimum > distance[vertex]) && (visited[vertex] == false) )
-        {
-            minimum = distance[vertex];
-            minVertex = vertex;
-        }
-    }
-    return minVertex;
-}
-
-void Dijkstra(graph, modified, src)
-{
-    var num_vertices = graph.length;
-    var distance = new List(num_vertices);
-
-    var visited = new List(num_vertices);
-
-    for (var i = 0; i < num_vertices; i++)
-    {
-        distance[i] = MAX_INT;
-        visited[i] = false;
-    }
-
-    distance[src] = 0;
-
-    for(var count = 0; count < num_vertices; count++)
-    {
-        var curVertex = minDistance(distance, visited);
-        visited[curVertex] = true;
-        for(var vertex = 0; vertex < num_vertices; vertex++)
-        {
-            if ((visited[vertex] == false) && (distance[vertex] > (distance[curVertex] +
-            modified[curVertex][vertex])) && (graph[curVertex][vertex] != 0))
-            {
-               distance[vertex] = (distance[curVertex] + modified[curVertex][vertex]);
-            }
-        }
-    }
-
-    for(var vertex = 0; vertex < num_vertices; vertex++)
-    {
-        print('Vertex   ${vertex} :  ${distance[vertex]}');
-    }
-}
-
-List BellmanFord(edges, graph, num_vertices)
-{
-    var distance = new List(num_vertices+1);
-
-    for(var i = 0; i <= num_vertices; i++)
-    {
-        distance[i]=MAX_INT;
-    }
-
-    distance[num_vertices] = 0;
-
-    for(var i = 0; i < num_vertices; i++)
-    {
-        edges.add([num_vertices, i, 0]);
-    }
-
-    for(var i = 0; i < num_vertices; i++)
-    {
-        for(var j in edges)
-        {
-
-            if((distance[j[0]] != MAX_INT) && (distance[j[0]] + j[2] < distance[j[1]]) )
-            {
-                distance[j[1]] = distance[j[0]] + j[2];
-            }
-        }
-    }
-
-    return distance;
-}
-
-
-void JohnsonAlgorithm(graph)
-{
-    var edges = new List();
-
-    for(var i = 0; i < graph.length ; i++)
-    {
-        for(var j = 0; j < graph[i].length; j++)
-        {
-          if(graph[i][j] != 0)
-          {
-              edges.add([i, j, graph[i][j]]);
-          }
-        }
-    }
-
-    var modifiedwei = BellmanFord(edges, graph, graph.length);
-
-    var modified = [[0,0,0,0],
-                        [0,0,0,0],
-                        [0,0,0,0],
-                        [0,0,0,0]];
-
-    for(var i = 0; i < graph.length; i++)
-    {
-        for(var j = 0; j < graph[i].length; j++)
-        {
-            if(graph[i][j] != 0)
-            {
-                modified[i][j] = (graph[i][j] +
-                modifiedwei[i] - modifiedwei[j]);
-            }
-        }
-    }
-
-    print ('Modified Graph: ${modified}');
-
-    for(var src = 0; src < graph.length; src++)
-    {
-        print ('\nShortest Distance with vertex ${src} as the source:\n');
-        Dijkstra(graph, modified, src);
-    }
-}
-
-void main() {
-
-    var graph = [[0, -8, 2, 4],
-                  [0, 0, 2, 6],
-                  [0, 0, 0, 2],
-                  [0, 0, 0, 0]];
-
-    JohnsonAlgorithm(graph);
-}
-
-/*
-Modified Graph: [[0, 0, 8, 8], [0, 0, 0, 2], [0, 0, 0, 0], [0, 0, 0, 0]]
-
-Shortest Distance with vertex 0 as the source:
-
-Vertex   0 :  0
-Vertex   1 :  0
-Vertex   2 :  0
-Vertex   3 :  0
-
-Shortest Distance with vertex 1 as the source:
-
-Vertex   0 :  9223372036854775807
-Vertex   1 :  0
-Vertex   2 :  0
-Vertex   3 :  0
-
-Shortest Distance with vertex 2 as the source:
-
-Vertex   0 :  9223372036854775807
-Vertex   1 :  9223372036854775807
-Vertex   2 :  0
-Vertex   3 :  0
-
-Shortest Distance with vertex 3 as the source:
-
-Vertex   0 :  9223372036854775807
-Vertex   1 :  9223372036854775807
-Vertex   2 :  -9223372036854775801
-Vertex   3 :  0
-
- */

From ac84dbda2122eaedca5277341e3fcb7f83608062 Mon Sep 17 00:00:00 2001
From: Yashasvi Sinha <59117774+heisastark@users.noreply.github.com>
Date: Fri, 3 Apr 2020 01:36:16 +0530
Subject: [PATCH 5/8] deleting accidentally added file

---
 .../Logistic_Regression.ipynb                 | 418 ------------------
 1 file changed, 418 deletions(-)
 delete mode 100644 Machine_Learning/Logistic_Regression/Logistic_Regression.ipynb

diff --git a/Machine_Learning/Logistic_Regression/Logistic_Regression.ipynb b/Machine_Learning/Logistic_Regression/Logistic_Regression.ipynb
deleted file mode 100644
index 669cfdb7fc..0000000000
--- a/Machine_Learning/Logistic_Regression/Logistic_Regression.ipynb
+++ /dev/null
@@ -1,418 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "code",
-   "execution_count": 1,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "import numpy as np\n",
-    "import matplotlib.pyplot as plt\n",
-    "import pandas as pd\n",
-    "%matplotlib inline"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 2,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "plt.style.use('seaborn')"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 3,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "(500, 2)\n",
-      "(500, 2)\n"
-     ]
-    }
-   ],
-   "source": [
-    "mean_01 = np.array([1,0.5])\n",
-    "cov_01 = np.array([[1,0.1],[0.1,1.2]])\n",
-    "\n",
-    "mean_02 = np.array([4,5])\n",
-    "cov_02 = np.array([[1.21,0.1],[0.1,1.3]])\n",
-    "\n",
-    "\n",
-    "# Normal Distribution\n",
-    "dist_01 = np.random.multivariate_normal(mean_01,cov_01,500)\n",
-    "dist_02 = np.random.multivariate_normal(mean_02,cov_02,500)\n",
-    "\n",
-    "print(dist_01.shape)\n",
-    "print(dist_02.shape)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 4,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 576x396 with 1 Axes>"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "plt.figure(0)\n",
-    "\n",
-    "plt.scatter(dist_01[:,0],dist_01[:,1],label='Class 0')\n",
-    "plt.scatter(dist_02[:,0],dist_02[:,1],color='r',marker='^',label='Class 1')\n",
-    "plt.xlim(-5,10)\n",
-    "plt.ylim(-5,10)\n",
-    "plt.xlabel('x1')\n",
-    "plt.ylabel('x2')\n",
-    "plt.legend()\n",
-    "plt.show()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 5,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "(500, 1)\n",
-      "(500, 1)\n",
-      "(1000, 3)\n"
-     ]
-    }
-   ],
-   "source": [
-    "y1 = np.zeros((500,1))\n",
-    "print(y1.shape)\n",
-    "\n",
-    "y2 = np.ones((500,1))\n",
-    "print(y2.shape)\n",
-    "\n",
-    "data1 = np.hstack((dist_01,y1))\n",
-    "data2 = np.hstack((dist_02,y2))\n",
-    "data = np.vstack((data1,data2))\n",
-    "\n",
-    "print(data.shape)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 6,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "[[ 4.61546313  2.66193635  1.        ]\n",
-      " [ 4.14312954  4.21854945  1.        ]\n",
-      " [ 3.5360558   5.44785463  1.        ]\n",
-      " [-0.30283189  1.49333376  0.        ]\n",
-      " [ 3.78745303  6.42600424  1.        ]\n",
-      " [ 1.88347777  4.75120101  1.        ]\n",
-      " [ 4.62331128  4.57041199  1.        ]\n",
-      " [ 3.25889833  6.95708425  1.        ]\n",
-      " [ 0.8891156  -1.34359916  0.        ]\n",
-      " [ 3.17329477  3.51864788  1.        ]]\n"
-     ]
-    }
-   ],
-   "source": [
-    "np.random.shuffle(data)\n",
-    "print(data[:10])"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 7,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "split = int(0.8*data.shape[0])"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 8,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "(800, 2) (200, 2)\n",
-      "(800,) (200,)\n"
-     ]
-    }
-   ],
-   "source": [
-    "X_train = data[:split,:-1]\n",
-    "X_test = data[split:,:-1]\n",
-    "\n",
-    "Y_train = data[:split,-1]\n",
-    "Y_test  = data[split:,-1]\n",
-    "\n",
-    "print(X_train.shape,X_test.shape)\n",
-    "print(Y_train.shape,Y_test.shape)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 9,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "def hypothesis(x,w,b):\n",
-    "    '''accepts input vector x, input weight vector w and bias b'''\n",
-    "    \n",
-    "    h = np.dot(x,w) + b\n",
-    "    return sigmoid(h)\n",
-    "\n",
-    "def sigmoid(z):\n",
-    "    return 1.0/(1.0 + np.exp(-1.0*z))\n",
-    "\n",
-    "def error(y_true,x,w,b):\n",
-    "    \n",
-    "    m = x.shape[0]\n",
-    "    \n",
-    "    err = 0.0\n",
-    "    \n",
-    "    for i in range(m):\n",
-    "        hx = hypothesis(x[i],w,b) \n",
-    "        err += y_true[i]*np.log2(hx) + (1-y_true[i])*np.log2(1-hx)\n",
-    "        \n",
-    "    \n",
-    "    return -err/m\n",
-    "\n",
-    "\n",
-    "def get_grads(y_true,x,w,b):\n",
-    "    \n",
-    "    grad_w = np.zeros(w.shape)\n",
-    "    grad_b = 0.0\n",
-    "    \n",
-    "    m = x.shape[0]\n",
-    "    \n",
-    "    for i in range(m):\n",
-    "        hx = hypothesis(x[i],w,b)\n",
-    "        \n",
-    "        grad_w += -1*(y_true[i] - hx)*x[i]\n",
-    "        grad_b +=  -1*(y_true[i]-hx)\n",
-    "        \n",
-    "    \n",
-    "    grad_w /= m\n",
-    "    grad_b /= m\n",
-    "    \n",
-    "    return [grad_w,grad_b] #returning python list\n",
-    "\n",
-    "\n",
-    "# One Iteration of Gradient Descent\n",
-    "def grad_descent(x,y_true,w,b,learning_rate=0.1):\n",
-    "    \n",
-    "    err = error(y_true,x,w,b)\n",
-    "    [grad_w,grad_b] = get_grads(y_true,x,w,b)\n",
-    "    \n",
-    "    w = w - learning_rate*grad_w\n",
-    "    b = b - learning_rate*grad_b\n",
-    "    \n",
-    "    return err,w,b\n",
-    "    \n",
-    "def predict(x,w,b):\n",
-    "    \n",
-    "    confidence = hypothesis(x,w,b)\n",
-    "    if confidence<0.5:\n",
-    "        return 0\n",
-    "    else:\n",
-    "        return 1\n",
-    "    \n",
-    "def accuracy(x_tst,y_tst,w,b):\n",
-    "    \n",
-    "    y_pred = []\n",
-    "    \n",
-    "    for i in range(y_tst.shape[0]):\n",
-    "        p = predict(x_tst[i],w,b)\n",
-    "        y_pred.append(p)\n",
-    "        \n",
-    "    y_pred = np.array(y_pred)\n",
-    "    \n",
-    "    return  float((y_pred==y_tst).sum())/y_tst.shape[0]\n",
-    "        \n",
-    "    \n",
-    "    "
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 10,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "loss = []\n",
-    "acc = []\n",
-    "\n",
-    "W = 2*np.random.random((X_train.shape[1],))\n",
-    "b = 5*np.random.random()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 11,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "for i in range(300):\n",
-    "    l,W,b = grad_descent(X_train,Y_train,W,b,learning_rate=0.5)\n",
-    "    acc.append(accuracy(X_test,Y_test,W,b))\n",
-    "    loss.append(l)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 12,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 576x396 with 1 Axes>"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "plt.plot(loss)\n",
-    "plt.ylabel(\"negative of log likelihood\")\n",
-    "plt.xlabel(\"epoch\")\n",
-    "plt.show()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 13,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 576x396 with 1 Axes>"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "0.98\n"
-     ]
-    }
-   ],
-   "source": [
-    "plt.plot(acc)\n",
-    "plt.ylabel(\"accuracy\")\n",
-    "plt.xlabel(\"epoch\")\n",
-    "plt.show()\n",
-    "print(acc[-1])"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 14,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 576x396 with 1 Axes>"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "plt.figure(0)\n",
-    "\n",
-    "plt.scatter(dist_01[:,0],dist_01[:,1],label='Class 0')\n",
-    "plt.scatter(dist_02[:,0],dist_02[:,1],color='r',marker='^',label='Class 1')\n",
-    "plt.xlim(-5,10)\n",
-    "plt.ylim(-5,10)\n",
-    "plt.xlabel('x1')\n",
-    "plt.ylabel('x2')\n",
-    "\n",
-    "x = np.linspace(-4,8,10)\n",
-    "y = -(W[0]*x + b)/W[1]\n",
-    "plt.plot(x,y,color='k')\n",
-    "\n",
-    "plt.legend()\n",
-    "plt.show()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 15,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "[0.76604691 1.49267387]\n",
-      "-5.424920453033558\n"
-     ]
-    }
-   ],
-   "source": [
-    "print(W)\n",
-    "print(b)"
-   ]
-  },
-  {
-   "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": "3.6.8"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 2
-}

From 536406b039309ae9f9e6c82980ffcc66039e69c0 Mon Sep 17 00:00:00 2001
From: Yashasvi Sinha <59117774+heisastark@users.noreply.github.com>
Date: Fri, 3 Apr 2020 01:36:25 +0530
Subject: [PATCH 6/8] deleting accidentally added file

---
 .../Tree_Preorder_Traversal.kt                | 50 -------------------
 1 file changed, 50 deletions(-)
 delete mode 100644 Tree_Preorder_Traversal/Tree_Preorder_Traversal.kt

diff --git a/Tree_Preorder_Traversal/Tree_Preorder_Traversal.kt b/Tree_Preorder_Traversal/Tree_Preorder_Traversal.kt
deleted file mode 100644
index 18bef0bd6c..0000000000
--- a/Tree_Preorder_Traversal/Tree_Preorder_Traversal.kt
+++ /dev/null
@@ -1,50 +0,0 @@
-import java.util.*
-class Node<Int>(
-    var key:Int,
-    var left:Node<Int>? = null,
-    var right:Node<Int>? = null  
-)
-{
-    fun preorderTraversal()
-    {
-      print("$key ")
-      left?.preorderTraversal()
-      right?.preorderTraversal()
-    }
-}
-
-fun main() {
-    var read = Scanner(System.`in`)
-    println("Enter the size of Array:")
-    val arrSize = read.nextLine().toInt()
-    var arr = IntArray(arrSize)
-    val nodes = mutableListOf<Node<Int>>()
-    println("Enter the array respresentaion of binary tree")
-    for(i in 0 until arrSize)
-    {
-        arr[i] = read.nextLine().toInt()
-        nodes.add(Node(arr[i]))
-    }
-    for(i in 0..arrSize-2)
-    {
-        if((i*2)+1<arrSize && arr[(i*2)+1]!=-1)
-        nodes[i].left = nodes[(i*2)+1]
-        if((i*2)+2<arrSize && arr[(i*2)+2]!=-1)
-        nodes[i].right = nodes[(i*2)+2]
-    }
-    print("Pre Order traversal of tree is ")
-    nodes[0].preorderTraversal()
-}
-/*
-*Enter the size of Array:
-*7
-*Enter the array respresentaion of binary tree
-*1
-*2
-*3
-*4
-*5
-*-1
-*7
-*Pre Order traversal of tree is 1 2 4 5 3 7 
-*/

From 3461111c6e332471388036f1291a1384d93a6b1e Mon Sep 17 00:00:00 2001
From: Yashasvi Sinha <59117774+heisastark@users.noreply.github.com>
Date: Fri, 3 Apr 2020 01:36:35 +0530
Subject: [PATCH 7/8] deleting accidentally added file

---
 kosaraju_algorithm/kosaraju_algorithm.cpp | 105 ----------------------
 1 file changed, 105 deletions(-)
 delete mode 100644 kosaraju_algorithm/kosaraju_algorithm.cpp

diff --git a/kosaraju_algorithm/kosaraju_algorithm.cpp b/kosaraju_algorithm/kosaraju_algorithm.cpp
deleted file mode 100644
index 8b969d461e..0000000000
--- a/kosaraju_algorithm/kosaraju_algorithm.cpp
+++ /dev/null
@@ -1,105 +0,0 @@
-//Strongly connected components(KOSARAJU ALGORITHM)
-#include <iostream>
-#include <vector>
-#include <stack>
-#include <unordered_set>
-using namespace std;
-
-//dfs to store vertices in a stack as per their finish time
-void dfs(vector<int>* edges, int start, unordered_set<int> &visited, stack<int> &finishStack) {
-    visited.insert(start);
-    for (int i = 0; i < edges[start].size(); i++) {
-        int adjacent = edges[start][i];
-        if (visited.count(adjacent) == 0) {
-            dfs(edges, adjacent, visited, finishStack);
-        }
-    }
-    finishStack.push(start);
-}
-
-//dfs of a transpose graph
-void dfs2(vector<int>* edges, int start, unordered_set<int>* component, unordered_set<int> & visited) {
-    visited.insert(start);
-    component->insert(start);
-    for (int i = 0; i < edges[start].size(); i++) {
-        int adjacent = edges[start][i];
-        if (visited.count(adjacent) == 0) {
-            dfs2(edges, adjacent, component, visited);
-        }
-    }
-}
-
-unordered_set<unordered_set<int>*>* getSCC(vector<int>* edges, vector<int>* edgesT, int n) {
-    unordered_set<int> visited;
-    stack<int> finishedVertices;
-    for (int i = 0; i < n; i++) {
-        if (visited.count(i) == 0) {
-            dfs(edges, i, visited, finishedVertices);
-        }
-    }
-    unordered_set<unordered_set<int>*>* output = new unordered_set<unordered_set<int>*>();
-    visited.clear();
-    while (finishedVertices.size() != 0) {
-        int element = finishedVertices.top();
-        finishedVertices.pop();
-        if (visited.count(element) != 0) {
-            continue;
-        }
-        unordered_set<int>* component = new unordered_set<int>();
-        dfs2(edgesT, element, component, visited);
-        output->insert(component);
-    }
-    return output;
-}
-
-int main() {
-    int n;    //no.of vertices
-    cout<<"Enter the no. of vertices : "<<endl;
-    cin >> n;
-    vector<int>* edges = new vector<int>[n];    //using a adjacency list to implement
-    vector<int>* edgesT = new vector<int>[n];    //transpose of a graph
-    int m;    // no.of edges
-    cout<<"Enter the no. of edges : "<<endl;
-    cin >> m;
-
-    for (int i = 0; i < m; i++) {
-        int j, k;
-        cin >> j >> k;
-        edges[j - 1].push_back(k - 1);
-        edgesT[k - 1].push_back(j - 1);
-    }
-
-    unordered_set<unordered_set<int>*>* components = getSCC(edges, edgesT, n);
-    unordered_set<unordered_set<int>*>::iterator it = components->begin();
-
-    while (it != components->end()) {
-        unordered_set<int>* component = *it;
-        unordered_set<int>::iterator it2 = component->begin();
-        while (it2 != component->end()) {
-            cout <<*it2  + 1<< " ";
-            it2++;
-        }
-        cout << endl;
-        delete component;
-        it++;
-    }
-
-    delete components;
-    delete [] edges;
-    delete [] edgesT;
-}
-
-/*INPUTS
-Enter the no. of vertices : 6
-Enter the no. of edges  :7
-1 2
-2 3
-3 4
-4 1
-3 5
-5 6
-6 5
-
-OUTPUT
-6 5
-2 3 4 1 */

From 30e214e8b2c71702de256ee2e8a9d4c49a295c19 Mon Sep 17 00:00:00 2001
From: Yashasvi Sinha <59117774+heisastark@users.noreply.github.com>
Date: Fri, 3 Apr 2020 02:37:46 +0530
Subject: [PATCH 8/8] Karatsuba_Algorithm.dart added (accidently deleted while
 squashing commits)

---
 Karatsuba_Algorithm/Karatsuba_Algorithm.dart | 57 ++++++++++++++++++++
 1 file changed, 57 insertions(+)
 create mode 100644 Karatsuba_Algorithm/Karatsuba_Algorithm.dart

diff --git a/Karatsuba_Algorithm/Karatsuba_Algorithm.dart b/Karatsuba_Algorithm/Karatsuba_Algorithm.dart
new file mode 100644
index 0000000000..6fa0b0b3cf
--- /dev/null
+++ b/Karatsuba_Algorithm/Karatsuba_Algorithm.dart
@@ -0,0 +1,57 @@
+import 'dart:io';
+import 'dart:math';
+
+// function to find minimun of two integers
+int minof(a, b){
+    if(a < b) 
+        return a;
+    else return b;
+}
+
+int karatSuba(num1, num2){
+
+    // for single digit number multiply directly
+    if (num1 < 10 || num2 < 10) {
+        return num1 * num2;
+    }
+
+    String num1Str = num1.toString();
+    String num2Str = num2.toString();
+    int n = minof(num1Str.length, num2Str.length);
+    int half = (n / 2).round();
+
+    // divide num1 into two halves
+    int num1_H = int.parse(num1Str.substring(0, num1Str.length - half));
+    int num1_L = int.parse(num1Str.substring(num1Str.length - half, num1Str.length));
+
+    // divide num2 into two halves
+    int num2_H = int.parse(num2Str.substring(0, num2Str.length - half));
+    int num2_L = int.parse(num2Str.substring(num2Str.length - half, num2Str.length));
+
+    // using the KaratSuba Definition
+    int s1 = karatSuba(num1_L, num2_L);
+    int s2 = karatSuba(num1_L + num1_H, num2_L + num2_H);
+    int s3 = karatSuba(num1_H, num2_H);
+    int s4 = s2 - s3 - s1;
+
+    // karatsuba formula for finding product
+    int result = s3 * pow(10, 2 * half) + s4 * pow(10, half) + s1;
+
+    return result;
+}
+
+void main(){
+    String a, b;
+    a = stdin.readLineSync(); 
+    b = stdin.readLineSync();
+    print(karatSuba(int.parse(a), int.parse(b)));
+}
+
+/*
+Sample input:
+12345
+6789
+
+Sample Output:
+83810205
+*/