Chatlytics is a data query and visualization platform. Part of the platform is a bot called "Fetch". Think of Fetch not just as a chatbot, but as an analytics assistant too.
Fetch has one key job to perform: help teams use their data to create beautiful visualizations without ever needing to leave Slack (or Hipchat). Fetch takes care of the query, visualization and reporting grunt work so your team can focus on more important things. Fetch packages everything in flexible and customizable PNG & interactive SVGs.
- Prerequisites
- How To Interact With Fetch, Your Analytics Bot
- Teaching Fetch Custom Commands
- Creating Your Own Themes
- Notes
- Issues
- Contributing
You will need:
- Database connection parameters. Currently only
PostgreSQL,MySQL,Amazon Redshift,Amazon Aurora (Postgres/MYSQL)andAmazON RDS (Postgres/MYSQL)are supported) - Slack account for the bot to use. If you do not have one you need to get yourself a Slack account. Go to the Slack website: https://slack.com/
The first step required to set the bot up involves clicking the "Add to Slack" button at https://www.chatlytics.co.
Chatlytics is only available for use with Slack at the moment. Other chat clients (Hipchat, Facebook, Skype, etc.) will be added soon!
You'll want to start by setting up your database connection. Run the command !dbconfig and follow the prompts to get started.
In addition to your database type, Fetch will need your host address (which can be an IP address or domain), your username & password, port and database name. After you provide this info, you'll be able to review the configuration and start giving Fetch commands!
Once you have connected Chatlytics to your database, you can interact with Fetch by sending commands to the linked Slack account.
You'll find a guide to these commands at https://www.chatlytics.co/guide.html. Just select a command in the dropdown to see what Fetch will do with it and get additional tips.
Fetch has a built-in set of commands it understands. To see a list of available commands Fetch can recognize, type !help.
The following is a list of available predefined commands:
-
!list_tables: Returns a list of all tables in the database. -
!table_count <table_name>: Returns the number of rows in the table.- Example:
!table_count MyTable
- Example:
-
!count_compare <table_name> <date_column> <style>: Generate gauge chart by comparing record count to the previous day.-
Example:
!count_compare MyTable timestamp blue
-
-
!date_compare <table_name> <date_column> <date1> <date2> <style>: Generate gauge chart by comparing record count of two different days.- Example:
!date_compare MyTable timestamp 01/01/16 07/07/16 sunset
- Example:
-
!chart <table_name> <date_column> <number_of_days> <style>: Generate a bar chart of record counts from the specified table for the given number of days.-
Example:
!chart MyTable timestamp 10 spring
-
Fetch will return charts in Slack, as well as a link to view an interactive SVG of the chart.
Here are some sample charts Fetch generated that you can interact with:
In addition to the predefined commands Fetch understands, you can teach it new ones. This is done by feeding Fetch YAML files.
How do you train Fetch to understand new commands? There are three easy steps:
- Define Your Command In YAML
- You Train Fetch To Understand The Commands
- You Ask Fetch To Run Your Commands
To teach Fetch a new command, you need to first create a YAML definition. This is what a sample YAML definition looks like:
queries:
- name: test_query
query: select testColumn, count(*) from testTable group by testColumn
chart_type: bar_chart
chart_description: test
style: blue
Each element serves a purpose in helping Fetch understand what you want to have happen:
name: The name of the query. This ultimately becomes the command you use to call the query once it's defined.query: The SQL query to run.chart_type: The type of chart which you wish to generate. Currently, valid inputs are "bar_chart", "line_chart" and "table".chart_description: This description will be included when your query is run.style: the name of the style which you want the output to use
When writing custom queries for bar_chart or line_chart graph types, it is expected that the first column will be the x-axis and the second column will be the y-axis.
With your YAML file completed, it is time to teach Fetch the new command. This will require that you import or "feed" the YAML file to Fetch. There are three methods for feeding Fetch new commands:
- Drag-and-Drop
- Attachments
- Remote Import from URL
The simplest way to import a command is by simply dragging a YAML file to the Slack window and clicking the "Upload" button. The bot will do the rest!
On success, you'll see this:
Alternatively, you can upload a YAML file by clicking the '+' sign at the bottom of the chat window, navigating to the file on your computer and clicking the "Upload" button.
To import a remote YAML file from someplace like Github, you must specify a URL to that YAML file in the following format:
!define <link_to_yaml>
For example, if the link is from GitHub you'll need to click the Raw button at the top of the file and use the resulting URL.
!define https://raw.githubusercontent.com/dtroberts/TestRepo/master/ymls/multidef.ymlOnce the query is defined and imported, it will be accessible via !name, where name is the value of the name field in the YAML file.
Here is an example of a YAML file which would create a command !my_first_command:
queries:
- name: my_first_command
query: select timestamp::date, sum(subscribers) from MyTable where datediff(day, end_time::date, getdate()::date) < 30 group by timestamp::date
chart_type: bar_chart
chart_description: My first query definition
style: blue
This would output a chart with the daily number of subscribers over the last 30 days.
It is also possible to define multiple commands in the same file. This YAML file would define the commands !MainTableQuery and !OtherTableQuery:
queries:
- name: MainTableQuery
query: select testColumn, count(*) from MainTable group by date
chart_type: bar_chart
chart_description: My first query definition
style: sunset
- name: OtherTableQuery
query: select testColumn, count(*) from OtherTable group by timestamp
chart_type: line_chart
chart_description: Line chart for OtherTable record counts
style: spring
IMPORTANT: Custom commands are removed upon shutting down or restarting servers. In a future release we will add persistence capabilities when training Fetch.
Did you know that Fetch allows you to customize the look and feel of your visualizations? You can fully customize the colors and typography to fit your own preferences. Themes define the style of the visualizations Fetch delivers to your team. Each Theme may be different, offering many choices to easily change the look and feel. Themes also separate the presentation styles from the underlying data. Using Themes allows you to quickly change the visual style and layout of a visualization with simple CSS edits.
You may wish to create Themes for your own use, for a client project or to be shared with the Chatlytics community!
Themes are simple to create and use. All you need is clean, valid CSS. Take a look at the theme how-to guide and templates.
MySQL connections cannot be opened in read-only mode. As such, any query defined here can be run on the database. The best solution to ensure the security and integrity of your database is to create a new user for use with Chatlytics with read-only permissions and specify those credentials in !dbconfig.
For instructions on how to set up a read-only user, see the guide.
If you have any problems with or questions about this image, please contact us through a GitHub issue.
You are invited to contribute new features, fixes, or updates, large or small; we are always thrilled to receive pull requests, and do our best to process them as fast as we can.
Before you start to code, we recommend discussing your plans through a GitHub issue, especially for more ambitious contributions. This gives other contributors a chance to point you in the right direction, give you feedback on your design, and help you find out if someone else is working on the same thing.