Quantopix Analytics System (QAS)

Getting Started

To get started, each user needs an account on QAS. You can create an account by filling out the registration form and clicking on Register.
QAS Registration form
Soon after submission, you will receive a verification email. You must click on the link in the email to activate your account. Once activated, you will be directed to a login screen.
login form
The initial page after login is your 'Dashboard'. The Dashboard will contain all the shortcuts to the objects you save from previous work on the system. The Dashboard is organized in sections for each type of object you add. Any object with a toggle button (+Dash) dashboard button can be added to the dashboard by a click. If an object is already on the dashboard, it will have a minus sign(-Dash) minus dashboard button. Clicking on the (-Dash) button will remove the object from the Dashboard and the sign will change back to plus(+). Objects already on the Dashboard can be removed simply by clicking on the (-Dash) button.

Main Menu

  • Go to Dashboard
  • The dashboard is a convenient screen for accessing your favorite data objects, charts, apps, and models. Objects added to the dashboard will remain on it until manually removed or the dashboard is Reset/Cleared.
  • Reset/Clear Dashboard
  • Clears the dashboard of all objects and resets it to default contents
    Data Sources
  • Connect to New Data Source
  • Connecting to a remote data source allows you to analyze the most up-to-date data or run queries for real-time monitoring and analysis applications.
    The first time you connect to a data source, the data source definition gets cataloged in your account and will remain cataloged until it is manually deleted using the Data Sources Management Panel. A data source can be in "Connected" or "Disconnected" state. Note that any application or model which references a data object from a disconnected data source will fail.
    Connecting to a new data source requires at minimum the following:
    1. The hostname or IP address of the server where the data source resides and it is accessible from the QAS Server
    2. The data source is configured to allow remote access
    3. The data source (or Database) name as configured at the remote server
    4. A valid userid and password to connect to the data source (configured at the remote data source)
    QAS uses ODBC "Connection String" format to establish connections to data sources. Connection Strings have a standard format which takes the form of "key1=value1;key2=value2, etc". For specific information about the connection string to your database, check your database documentation or with your database administrator. An example of a connection string to Oracle database would look like this:
     "driver=Oracle; dbq=//myHostName:1521/myOracleDBName; uid=myUserID; 
       pwd=myPassword; description=Oracle Database on myHostName;" 

    For IBM DB2, the connection string may look like this:
     "driver=DB2; server=localhost; description=IBM DB2 on myServer; 
    	 database=SAMPLE; uid=myDB2UserID; pwd=myDB2Password; " 

    For MySQL PDO Connection, the string will be like this:
     "driver=PDO; server=myHostName; description=MySQL Database; 
    	 database=myDatabaseName; uid=myUserID; pwd=myPassword;  " 

    QAS allows you to generate the connection string for a supported data source through an input form. For example, to generate the connection string to Oracle database, select "Oracle ODBC Native Driver" from the drop-down menu
    Connect to database drop down
    then fill out the form
    Connect to oracle form
    Test the connection by clicking on the "Test Connection" Button
    Test Connection
    If the connection succeeded, you must click the "Save Connection" button to catalog it in your account. If the connection fails, you should see an error message describing the failure.
    Every new user on QAS has two built-in private databases created at first login:
    MyUploads: This database is intended for your data uploads.
    SampleData: This database contains tables used by the sample Apps that come with the system.
    Although these databases are not sharable with other users on the system, they are fully functional and can be used by your applications to create, store, and retrieve data.
  • Manage Data Sources
  • You can Connect, Disconnect, and Edit connection strings to cataloged data sources using this management panel
    Manage Data Sources
  • System Defined Data Sources
  • This option lists the pre-configured data sources for all users on the QAS server. On Linux, the configuration file cataloging these data sources must have the name
    and must be placed in the system directory
    All users who log into the QAS server will see and have access to these pre-configured data sources.
  • User Defined Data Sources
  • This option lists the data sources configured by the QAS users only. Other users on the QAS server cannot see or access these data sources.
  • Show ALL Data Sources
  • This option lists all data sources available for the user whether configured at the system level or by the user only.
    Data Objects
  • Online Tables
  • This option lists all Tables available to the user from all connected data sources. Tables from Data sources that are not connected will not show on this list
  • User Datasets
  • List Datasets
  • This option lists all datasets created by the user
  • Create New Dataset From SQL Query
  • This option starts the steps for the user to create a dataset from one of the tables online. To create a dataset, select the data source where the table resides then select the table from which you want to generate a dataset. You will be presented with a panel to build SQL query for a new dataset:
    New Dataset
    This form is intended to assist you in creating an SQL SELECT query from the selected table. If you are comfortable with SQL language and can write an SQL query without the form, then you have the option of doing so by switching to a free form input box. You can switch by clicking on Switch to SQL Free Form. When done building the SQL query for the new dataset, you can test it by clicking on Run query button button. Examine the results displayed below the form for problems and correct the query until the results are acceptable. Save the dataset by clicking on Save Query Button button. Now you have created a dataset which you can use as a data object in your analysis and applications. You can view the dataset by clicking on the "List Datasets" menu option or by clicking on option in the left pane of the page.
  • Create New Dataset From Data Filter
  • You don't need to know SQL to build the query needed for a dataset. Even if you do, sometimes it is not readily obvious what SQL query you need to filter the data to fit your analysis.
    QAS comes with a powerful filtering tool which allows you to experiment with a multitude of constraints and sorting logic to extract exactly the data you need.
    To generate a dataset using a data filter, select the data source where the table is, then select the table from which you want to generate a dataset. You will see the filtering form:  Data Filter
    Using the data filter is straight forward. Anytime you change the filtering parameters, the "Re-Generate Query" button turns red Regenerate Red Button and must be clicked to capture the changes. The resulting query will be displayed at the "Current Query" line .
    Finally, to turn the generated query into a dataset, you must save it by clicking Save Query as Dataset. You need to give your dataset a name and optionally a description: Save query from filter.
    Click on and your new dataset will be created.
  • Edit a Dataset
  • Editing an existing dataset allows you to change the description of the dataset and its SQL Query. To start, select a dataset to edit from the drop-down menu. In the edit form, modify the description and/or the SQL statement then save.
  • User Data Groups
  • A data group is a collection of Single Column Datasets. A single column dataset is simply a dataset made out of one column from any available Table on the system. Prior to creating a data group, you need to select which single columns it is made out of.
    QAS provides you with a quick way to create a single column dataset if you don't require filtering. Look for the "Make Dataset from xxxxxxx Column" in the Table or Dataset main page:
    Make Single Column Dataset
    Clicking on this button prompts you to accept or modify a system generated name for the dataset.
    Make single column prompt
    Click on Save Query to save it. Follow the same steps for every single column you want included in the data group.
    When you have all the datasets created, you are ready to build a data group.
  • List Data Groups
  • This option lists all existing data groups created by the user
  • Create New Data Group
  • This option starts the steps to creating a data group. First choose a name for the new group, enter it in the form below and submit:
    New data group name
    If you followed the steps above for creating "Single Column Datasets", you will see all the datasets eligible to be included in the new group listed in the left pane of the form.
    Data Group Form
    Highlight the datasets you like included in the data group (For multiple datasets, hold down the Ctrl key). Add the highlighted datasets to the right pane by clicking "Add Items" button.
    When done adding datasets, you must save the new data group by clicking "Save" at the bottom.
  • Edit a Data Group
  • This option lets you edit the contents of an existing data group. You will see the same form for creating a new group with the constituent datasets listed in the right pane of the form. You can Remove or Add datasets to the data group using the buttons between the two panes.
    To delete a data group, you simply remove all its datasets items. When the right pane is empty, click on Save and you will be prompted to delete the data group.

    Click "Delete Data Group" and the data group will be removed from your account.
  • Download Data From Table/Dataset
  • This option prompts you to download an existing Table or a Dataset to your computer. The data will be in Comma Separated Values (CSV) format. The CSV file will download as a zip file (*.zip).
    To start downloading, select the data source name of the Table or the Dataset. A drop-down list of all the available Tables and Datasets will be displayed. Select the object you want to download and click "Start Download"
    Download Data Object
  • Upload CSV File into Data Source
  • You can upload a Comma Separated Values (CSV) file to a new table in your account. The CSV file can be located on your hard drive or on a URL accessible to the QAS server on the internet. When you click on this menu option, you will be prompted for the source of the CSV file.
    Upload CSV File
    QAS supports *.csv, *.txt, and *.zip file extensions. If the extension of the uploaded file is *.zip, QAS will assume the content of the file is compressed and will attempt to unzip it before storing it.
    Note that uploading a data file to QAS happens in two steps: Uploading to the QAS server and Defining the target Table to receive the data on the system.
    Complete ALL the fields in the form and click "upload". This transfers the data file to the server but does not load it into a table. The uploaded file will not be accessible to you until you complete the second step.
    If the upload completes successfully, you will see QAS parsing of the uploaded file:
    Upload and Store CSV file
    Ensure that the column headers are filled out with valid column names and contain only alphanumeric characters.
    It is also very important to review and modify the column data type of each column. Data types determine which analysis can be performed on the column by QAS functions. Select the appropriate data type for each column and click the "LOAD IT" button.
    QAS will kick-off the process of loading the file into a data source table. This process takes place in the background on the server and does not impact your interface to QAS. You can check if the upload has completed successfully by examining the Tables in your target data source.
    Classification & Reports
    A classifier is a column in a data table that contains values used in multiple rows throughout the table. The column with the highest repeat count of a given value in the table rows is the column with the highest class or rank.
    QAS Classification is a way to represent Text and Integer data hierarchy for a given data object.
    Applying the classification to a data set will generate a hierarchical order of data classifiers for Text and Integer data types. A data column will have a higher class if it contains fewer unique values. Each unique value in a given column will serve as a pivot (a classifier) for the lower class columns. This allows the user to view the data set as a hierarchy of data dependencies. The algorithm will also generate a set of queries to extract and visualize unclassified data. Since classifiers provide pivots for the lower order data (more dependent data), there will be more permutations of queries generated for the lower order data and more charts to examine
  • Classify & Profile a Data Set
  • This option generates text classifications for the selected data object.
  • Create New Report
  • This option generates a report made of charts for all the columns in a given Table or Dataset. The charts are generated in the hierarchical order of the data classifier in the object. Report
    The example report below classifies the data in a Deck Of Cards table:
  • Show Generated Reports
  • This options displays the list of generated reports, their completion status, and options to View, Re-Generate, or Delete them from your account.
    Data Tools
  • Data Explorer
  • Data Explorer is a top-down visual exploration tool of Data Sources, Data Objects (Tables, User Defined Datasets, User Defined Data Groups), Apps, Reports, and Models. Check the left pane on the page for the list of objects you can click on to explore.
  • Data Visualizer
  • Data Visualizer is an interactive charting tool that allows you to plot the numeric columns in Tables, User Defined Datasets, and User Defined Data Groups.
  • Data Filter
  • This is a data filtering tool that allows you to experiment with different conditions for SQL queries while generating charts and tabulated results. You can change modify the value constraints for each column, sort row order by column in ascending or descending order, randomize row order, or limit the row count of the results. You can produce image charts or interactive charts on the fly while modifying the filtering parameters.
  • Copy User Dataset into New Table
  • This menu option can be used to create a new table from a user defined dataset. Select the dataset you want to copy, the target data source, then enter the name of the target table.
  • Copy User Data Group into New Table
  • his menu option can be used to create a new table from a user defined data group. Select the data group you want to copy, the target data source, then enter the name of the target table.
  • Global Text Search
  • Global Text Search allows you to search all online tables for a keyword
    Predictive Models
  • List Models
  • List all user created models
  • New Model
  • Multiple Linear Regression Model
  • This menu option brings up the form for generating a linear model from an existing data object.
  • Clustering Model (standard k-means)
  • This menu option brings up the form for generating a k-means clustering model from an existing data object
  • Predict
  • If you have generated predictive models, this option prompts the user for the input fields needed to make a prediction using the selected predictive model.
  • List Apps
  • This option lists all apps available to the account.
  • Create New App
  • Invokes the programming editor for a new App.
    When you open a new app, you start by giving the app a name and description:

    This will open the editor with a new app template. The template is meant to get your application started by including all objects and data structures useful for accessing data on the system and performing any type of analysis.
    You can elect to restrict your application to run only in a certain data context by editing the execution context definition
    The default for a new app is to have no context restriction which means that it will run whether data is available to it or not.
    Most applications require a data context to run, so selecting "Any Object" and then checking the type of object programmatically is a reasonable approach if your program requires data to run.
  • Edit Apps
  • Invokes the programming editor for an existing App
  • Turn Trace On/Off
  • Toggles user account level trace. Use this option to help you diagnose a failure in any of the system operations or in a user app.
    Caution: Turning trace ON can generate a large trace file if left on during normal operations. This can consume disk space and slow down the system for you and possibly for other users on the system. If you are debugging a problem, make sure you turn trace off once you capture the failure diagnosis.
  • View User Log File
  • Opens a new browser window to view the user trace file.
  • Clear User Log
  • clears the user trace file.

    Quantopix Analytics System (QAS) is published by
    Quantopix Technologies, LLC

    All rights reserved Quantopix © 2014