Skip to content
master
Switch branches/tags
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
CRM
 
 
 
 
css
 
 
 
 
js
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

CiviCRM Data Visualization Framework

Civisualize is a CiviCRM extension which let you visualize your CiviCRM data with the power of dc.js and d3.js. dc.js let users create graphs which enable filtering through the data with just a click on the graphs giving you a lot of powers. It's in beta, not going to hurt your data, do let us know if something works or doesn't.

Civisualize will help you out with ready to use visualizations that we have created and will provide you the right resources if you plan to create your own custom visualization.


Installation

Civisualize can be installed directly from within the CiviCRM user interface. See the user manual for details: Installing extensions.

If you are a developer you can clone Civisualize into your CiviCRM extension directory using:

git clone https://github.com/TechToThePeople/civisualize.git

Ready to Use Visualizations

The list of core visualizations is available at Reports > Civisualize. The currently available visualizations include:

Contributions (/dataviz/contribute)

alt tag This provides an overview of contributions, type of contributors, the day of week the contribution was made, payment instrument, etc.

Contacts (/dataviz/contacts)

alt tag Contacts overview is a compilation of graphs showing types, gender, age, source of contact, contacts over the time etc.

Events (/dataviz/events)

alt tag Events shows a lot of information about all the events of an organization and their participants. It shows events over time, participants over time, money generated from events, type of events, etc. It also displays a table listing all the events.

This table further links to specific event overview (/dataviz/event/) showing participants, their status, their fee etc.

alt tag

Donor Trends (/dataviz/donortrends)

alt tag Based on the idea of CiviCRM Donor Trends Extension, the donor trends overview showcases a Bar Chart with new, lapsed, upgraded, downgraded and maintained donors over the years. This further showcases the gender and age of the donors and a list of all the donors.

Need not be mentioned, you can add more graphs and details to the above mentioned visualizations.

Create your own visualizations

If you are a developer you can customize or create your extensions with a little knowledge of mysql/crmAPI, d3.js/dc.js and crossfilter. Civisualize basically has two parts:

  • Extract data from CiviCRM
  • Display data using dc.js/d3

Extract data

For the first part, we are using the following methods:

{crmAPI}

{crmAPI entity="OptionValue" option_group_id="14"}

You can add getstat as a new action on some entities in the api.

{crmSQL}

We have added a new crmSQL to run a mySQL query. For obvious reasons crmSQL only let you run SELECT queries. You can get your data using any of the following three methods:

  • SQL String

    {crmSQL sql="SELECT count(*) ... group by ...."}

  • SQL File

    {crmSQL query="somethingcool"}

    This will fetch the sql query from /queries/somethingcool.sql

  • JSON Object

    {crmSQL json="somethingcooler" cid=4 bla="hello"}

    This will fetch a json object from /queries/somethingcooler.json The format of the json is

    {  "query":"SELECT * from ABC where id=%1 and bla=%2",
       "params":{  "1":{ "name":"cid",
                         "type":"Integer"},
                   "2":{ "name":"bla",
                         "type":"String"}
                }
    }

    You can further use {crmRetrieve var="a" name="b" type="Integer"}. This will assign the POST or GET variable named b into a which can then be given to {crmSQL}.

    Optional Arguments

    {crmSQL set="varname"} will assign the result of the sql query to a smarty variable named varname.

    Return Value

    {crmSQL} returns a json of the following format:

       {"is_error":0,"error":"error_str","values":"Array of objects"}

    so we primarily use {crmSQL} values for our visualizations.

  • {crmReport}

    A 3rd option is to be able to fetch data from a report instance using {crmReport...}. Eileen has done (most of?) the work already. I think it's on 4.5.

Display data

The principle is to get the data in a template as a json, and apply dc on it until it looks awesome. You simply have to create a template into templates/dataviz/Something.tpl. Once you have the data from the above methods you can apply dc on it, and you can access it from /civicrm/dataviz/something.

To get you started, you can visit /civicrm/dataviz/contribute or any of the above mentioned visualizations.

This is using the wonderfully magic dc, that is a layer of love on the top of d3 and crossfilter. Click on the graphs to filter down. magic, I told you. No matter if you use {crmAPI} or {crmSQL}, you end up with a json and a d3 and dc loaded and ready to rock.

In the template, put:

<div id="theplacetograph"></div>
<script>
  // Load the data into 'mydata' from Smarty.
  var mydata={crmAPI or crmSQL};
  // Turn off Smarty's interpreter from here, it gets confused by Javascript and CSS.
  {literal}
  // We do everything in an anonymous function so we are free to use any
  // variable names we need without clobbering those in the global scope.
  (function() {
    // Define the function that will make your charts etc.
    function bootViz() {
      // Ensure we're using the up-to-date libraries, not the CiviCRM core ones.
      var d3 = CRM.civisualize.d3, dc = CRM.civisualize.dc, crossfilter = CRM.civisualize.crossfilter;

      //
      // This is where you do your stuff.
      //
      d3("#theplacetograph").selectAll(...).data(mydata.values).domagic(...);
    }

    // Boot our script as soon as ready.
    CRM.civisualizeQueue = CRM.civisualizeQueue || [];
    CRM.civisualizeQueue.push(bootViz);

  })(); // Immediately call our anonymous function.

We have also used {crmTitle} function which let you set the title of the page, and a print_filter function that will help you in playing around with crossfilter.

Check our resources at dc.js and create your first visualization.

Reset links

You can include reset links in your charts to reset all the filters for a particular chart.

  1. after defining your chart, save a reference to it using a unique name like so:
    CRM.civisualize.charts.yourChartsUniqueName = yourPieChart;

  2. At the end of these, call CRM.civisualize.bindResetLinks(); to make sure all the reset links are ready.

  3. Include links in your chart's container div like so:
    <a class="reset civisualize-reset" href data-chart-name="yourChartsUniqueName" >Reset</a>

The links should normally be hidden until you click to add a filter. See Contacts.tpl for an example use.

Documentation and help

If you want to add a document, add a markdown file that has the same name as the template into the doc folder.

We love you

xavier made this, sid helped him. You can find us on CiviCRM forum, @eucampaign and @sid0_o on Twitter, or in the dc mailing group. Be warned, d3 is awesome, but the learning curve is steep. Worthwhile, the view at the top is beautiful.

Changes

Version 6.1

  • Update civix boilerplate to work with PHP 7.4 without warnings.

Version 6.0 brings changes that will probably break your custom visualisations.

It became apparent that version 5's library updates conflicted with CiviCRM's core libraries. To avoid this v6 now puts its own versions of dc, d3 and crossfilter under CRM.civisualize as a namespace.

There were also loading problems: we need the DOM loaded (not as simple as it sounds, given the three types: normal page loads, cached dashboard, refreshed/new dashboard) and we need CRM.civisualize to have been created already, which it might not. To fix this we've implemented a queue similar to Google analytics etc.

And some issues with scripts using the global namespace for their vars.

And we tidied up the reset links.

So whereas before you might have just put your code in a <script> block and started calling d3.... now you should wrap your code and boot it as shown above.

Note you might find this breaks your dashboards. If this happens you'll need to clear your browser's "local storage".

Version 5.0 brings changes that will probably break your custom visualisations.

The libraries have been upgraded as follows:

  • d3 from 3.5.17 to 5.7.0
  • dc from 3.0.10 to 2.0.2
  • crossfilter to 1.3.12

The d3 upgrade requires changes to templates. These are detailed at https://github.com/d3/d3/blob/master/CHANGES.md.

There's a PHP command line helper tool updateTpl which may help... or not. It doesn't cover everything (feel free to improve it!), but it was able to convert all the bundled visualisations. Top tip: if/when you run it and you get a warning about d3.time.format().parse, it's best to revert to the backup (the script makes a backup for you), edit your original to a format that the converter can then handle. Once you've understood it, it's not difficult. Then run the converter and test.