Speaker 1: What are the tools in big data analytics? Which tool is the best tool among all? How it will help organization in analyzing the data? These are the questions you may arrive with. By watching this video, you will be able to understand all this. Hi, this is Sahana from Simply Learn. Today, we will learn which is the best tool for big data analytics. Before that, please make sure you subscribe to our channel and press the bell icon to never miss an update. Let's go for today's agenda. First, let's get introduced to what is big data, applications of big data. First comes Hadoop, after that Tableau, Cassandra, MongoDB and lastly Spark. And let's get started with what is big data. Big data can be defined as the large amount of data that is very hard to analyze using traditional data handling system like Microsoft Excel. Data generated through new technologies like smartphones, self-driving cars, social media websites like Facebook, Instagram and computers are categorized under big data. Such data sets are very hard to process using normal computing techniques. Next comes applications of big data. Big data is used in banking sector to analyze complex data sets of customers. Big data markets will analyze and anticipate the customer's liking while maintaining their expenditure record. This will happen in e-commerce websites like Amazon, Flipkart, Myntra, etc. Next, let us go for best big data tools. Most crucial tool is Hadoop. Hadoop is an open source tool written in Java that allows user to process large amount of data. Hadoop uses a network of computers to solve problems involving large amount of data. Hadoop is the best efficient tool to store large amount of data sets. Next let's go for features of Hadoop. Most important one is it is cost efficient because Hadoop runs on inexpensive commodity servers with associated storage, a less expensive design than a dedicated storage area network. Next is swiftness, data localization or the practice of performing computation close to the data rather than transporting the data itself. That is why Hadoop is very fast. Flexibility, Hadoop is extremely adaptable in terms of its capacity to handle various types of data sets like it may be structured, semi-structured, unstructured, all that can be handled efficiently using Hadoop. Scalability, a very scalable model is Hadoop. In a cluster, a sizable volume of data is split among several affordable processes and processed in simultaneously. Next let's understand the components of Hadoop. There are mainly three major components among Hadoop. First among them is Hadoop HDFS which is also called as Hadoop Distributed File System. Next comes MapReduce and the third is Hadoop YARN. Let's understand them in detail. First is HDFS. Based on the Google File System, the Hadoop Distributed File System offers a distributed file system that is intended to function on common hardware. It and current Distributed File System share a lot of similarities. The differences between this Distributed File System and others, however, are substantial. Hadoop is used in big organizations for both research and production. Hadoop also allows the transfer of data between the nodes. These nodes can also be addressed as NameNode and DataNode. Next comes MapReduce. MapReduce is a processing technique used in Hadoop for distributing computing with Java as a programming language. Important task of MapReduce is Map and Reduce. Map takes a set of datasets and converts them into another set, whereas Reduce converts them into small set of tuples. Next comes the third important component that is YARN. Important concept of YARN is to separate the resource management operations. The concept of YARN is to have an overall resource manager, which is shortly addressed as RM, and an application master for each application, which is shortly addressed as AM. Either a single job or a group of jobs make up an application. This is all about components of Hadoop. Now let's go for data storage in Hadoop. They are in the form of racks. First is rack 1, next comes rack 2, rack 3 and rack 4. Let's see how data is analyzed in Hadoop. Here block A is splitted as block A and another block A in rack 2 and block B is splitted in block 3. Same case for block C as well as block D and E. This is how data storage is very simple in Hadoop. Next comes crucial big data tool called Tableau. Tableau is one that tool that aids in maintaining a data pool. They have 40 distinct data sources that can have link to. Tableau software is an American interactive data visualization company, focuses on business intelligence. To visualize data while conducting business intelligence analysis, organizations employ Tableau. Tableau's product range consists of Visible, Tableau Public, Tableau Server, Tableau Desktop, Tableau Online and Tableau Mobile. It is used to visualize the data and explore different views. Tableau is applied in analyzing high volumes of data. Next let's go in detail about features of Tableau. Tableau Dashboard. Tableau Dashboards use text, graphic objects, visualizations and other elements to give you a complete picture of your data. Dashboards can show data as stories, allow the addition of various views and objects. It also offers a range of layouts and formats and allows users to apply appropriate features. These features make dashboards particularly informative. Data Sharing Among Tableau. Tableau offers simple way for users to work together and rapidly share the data in the form of visualizations, sheets, dashboards etc. You can use it to safely share the data from many different sources including hybrid, on-premise and on-cloud. Tableau Provides Advanced Visualization Technique for Detailed Visualization of Data. Data and User Security are given great considerations by Tableau. For data communications and use cases, it is a foolproof security system based on authentication and permission methods. These are all some of the important features provided by Tableau. This is a simple glimpse of Tableau Dashboard. Next comes Cassandra. Apache Cassandra is an open source tool which is having NoSQL database that is used for handling big data. Apache Cassandra has the capability to handle structured, semi-structured and unstructured data. It allows for storage and retrieving of data. Apache Cassandra was originally developed at Facebook, after that it was open sourced in 2008. Cassandra is an Apache product. It is an open source distributed and decentralized storage system. It is used to manage very large amount of structured data spread out across the world. It provides high availability with no single point of failure. Next is of Cassandra. Cassandra is a very highly scalable tool and allows more hardware utility. Cassandra provides flexible data storage and it supports all possible data formats like structured, semi-structured as well as unstructured data. Cassandra provides simple data distribution. It is very simple and also provide replication factor. Operations or table operations in Cassandra. First is creating a table. Command used to create a table i.e. create table and emp1 is the name of the table which has data sets like id and name of the employee. In Cassandra you can also drop a table like any other SQL language. It includes a simple command drop table emp1. Next is altering a table. It can be done using simple command alter table and table name i.e. emp1. Next is truncating a table. To do this operation you have to select the table and truncate the table by using the command truncate emp1. These are all the important or key table operations performed in Cassandra. This is how Cassandra will look like. Next comes very famous NoSQL database tool which is called as MongoDB. MongoDB is a document based, versatile and scalable NoSQL database management system that uses key value sets to store data and it also supports many data models. It was created as a way to deal with enormous amount of dispersed data that relational database or relational data models which typically provides rows, tables that cannot handle well. MongoDB is free and open source just like Hadoop. Features of MongoDB Data replication is the key feature of MongoDB. It creates replica datasets to enable fault tolerance. Data is kept on many servers using replication which offers high availability and redundancy. Flexible data storage like Cassandra and Hadoop it can also store any type of datasets like semi-structured, unstructured and structured. Next comes storage engine. It employs multiple storage engines thereby ensuring the right engine is used for the right workload which in turn enhances performance. It is a powerful query language that enables CRUD operations which stands for create, read, update and delete operations, text search and aggregation functions. Due to embedded data models, it requires less input and output procedures than relational databases. Faster queries are also supported by MongoDB indexes. These are all the key features of MongoDB. This is a glimpse of MongoDB Compass. Next comes Spark. Fast cluster computing called Apache Spark is made for quick computation. It was constructed on top of Hadoop MapReduce and expands the MapReduce concept to utilize other functions like calculations including interactive queries and stream processing. Spark is a very flexible and easy to use tool. Let's go and understand main important components of Spark. First among them is Spark Shell. Spark offers an interactive shell, a potent tool for interactive data analysis. You can get it in either Python or Scala language. A distributed collection of items known as Resilient Distributed Datasets that is RDD. RDD can be produced by altering existing RDDs or by using Hadoop input formats which is like HDFC files. RDD Transformations. You can establish dependencies between RDDs by using RDD Transformation which returns a pointer to a new RDD. In a dependency chain, each RDD includes a function for computing its data as well as dependency to its parent RDD. Spark Code. The basic functionality of Spark is carried out via Spark Code. It contains the parts necessary for managing memory, interacting with the storage system, recovering from errors and scheduling tasks. Next comes Spark SQL. On top of Spark Code, Spark SQL is constructed. Support for structured data is also offered. Both the SQL or Structured Query Language and HQL. Apache Hide variation of SQL are supported for data querying. Next comes Machine Learning Library. A machine learning library called MLIB or MLib includes a variety of machine learning algorithms. These comprise Principal Component Analysis, Classification, Regression, Clustering, Correlations and Hypothesis Testing. Compared to Apache Monads, it is 9 times quicker. Spark Stages. This is a glimpse of Spark Stages. Source is spark.org. These are all the 5 best tools for Big Data Analytics in 2022. Thank you for watching the video. In case of any queries, leave it in comment section. Thank you. you