Ahh, I see the problem now. You have a closure issue in play: by the time
the view is calculated, the "name" variable will contain the name of the
last entry, thus you will never get it to match anything else. Try this
instead:
for (var i = 0; i < columnNames.length; i++) {
name = columnNames[i];
columns.push({
type:'number',
label: name,
calc: (function (myName) {
return function (data, row) {
return (data.getValue(row, 1) == myName) ? data.getValue(row
, 2) : null;
}
})(name)
});
}
This locks the value of "name" to the "myName" variable inside the closure
so that when the returned function is executed, it has a unique value of
"myName".
On Friday, November 16, 2012 12:32:07 PM UTC-5, Iain wrote:
>
> Hi, very good of you have a look and reply.
>
> The variable data is the DataTable, and when I create the view below,
> you'll see I actually create it from another view of that DataTable called
> 'year_country_group'. In the DataTable, the countries column is column 2 -
> in the year_country_group view, the countries column is column 1.
>
> Thanks,
> Iain
>
>
>
> On Friday, November 16, 2012 4:56:00 PM UTC, asgallant wrote:
>>
>> At a guess, I'd say the problem is this line:
>>
>> var columnNames = data.getDistinctValues(2);
>>
>> You're getting the values from column 2, but comparing them to column 1
>> in the loop. I think you meant to have this:
>>
>> var columnNames = data.getDistinctValues(1);
>>
>> On Friday, November 16, 2012 9:28:02 AM UTC-5, Iain wrote:
>>>
>>> Darn, I thought the automation would be simple, but I have an weird
>>> error I can't figure out. The code below, in which the for loop automates
>>> what is then commented out, doesn't work. If I bring the comment into play
>>> to replace the automation however, I can go on to group the data to
>>> complete the pivot and draw a table or chart fine. But as far as I can
>>> see, the 'columns' array from either implementation is identical to the
>>> other. Any ideas?
>>>
>>> var columns = [0];
>>> var columnNames = data.getDistinctValues(2);
>>> for (var i=0; i<columnNames.length; i++) {
>>> name = columnNames[i];
>>> columns.push({type:'number', label: name, calc: function (data,
>>> row) {return (data.getValue(row, 1) == name) ? data.getValue(row, 2) :
>>> null;}});
>>> }
>>> /*
>>> var columns = [0,
>>> {type: 'number', label: 'Arg', calc: function (data, row) {return
>>> (data.getValue(row, 1) == 'Arg') ? data.getValue(row, 2) : null;}},
>>> {type: 'number', label: 'Bol', calc: function (data, row) {return
>>> (data.getValue(row, 1) == 'Bol') ? data.getValue(row, 2) : null;}},
>>> {type: 'number', label: 'Chi', calc: function (data, row) {return
>>> (data.getValue(row, 1) == 'Chi') ? data.getValue(row, 2) : null;}}
>>> ]
>>> */
>>> var view = new google.visualization.DataView(year_country_grouping);
>>> view.setColumns(columns);
>>>
>>> Thanks again,
>>> Iain
>>>
>>> On Friday, November 16, 2012 12:07:13 PM UTC, Iain wrote:
>>>>
>>>> Nice - I've written an implementation of that that works - now I just
>>>> have to figure out how to automate it as I won't know what the data will
>>>> be
>>>> - some form of loop through the distinct values of the given column I
>>>> expect.
>>>>
>>>> Any experience of the performance of this pivot function for real data
>>>> sets?
>>>>
>>>> Thanks for your help, really appreciated.
>>>> Iain
>>>>
>>>> On Thursday, November 15, 2012 8:25:52 PM UTC, asgallant wrote:
>>>>>
>>>>> You need to pivot your data, so you end up with one data series for
>>>>> each country. There isn't any support for pivots in the API, but I wrote
>>>>> a
>>>>> hack that shows you how to do a pivot manually:
>>>>> http://jsfiddle.net/asgallant/HkjDe/
>>>>>
>>>>> On Thursday, November 15, 2012 1:45:18 PM UTC-5, Iain wrote:
>>>>>>
>>>>>> Hi, hoping to get some help with DataViews. My ultimate goal is to
>>>>>> draw Pie Charts and Column Charts form one DataTable - my understanding
>>>>>> is
>>>>>> that I can create different DataViews from that DataTable to support
>>>>>> this.
>>>>>>
>>>>>> I've created a DataTable with columns 'KGs', 'Year', 'Country', and
>>>>>> 'Material'. I was able to create a pie chart that aggregated each
>>>>>> country's KG's by year using the following code:
>>>>>>
>>>>>> var by_year = google.visualization.data.group(dt, [1], [{'column': 0,
>>>>>> 'aggregation': google.visualization.data.sum, 'type': 'number'}]);
>>>>>> var chart = new
>>>>>> google.visualization.PieChart(document.getElementById('chart_year'));
>>>>>> chart.draw(by_year, options);
>>>>>>
>>>>>> I then wanted to create a Column Chart with KGs on the y-axis, Year
>>>>>> on the x-axis and Country in the Legend. I grouped the data as follows:
>>>>>>
>>>>>> var by_year_country = google.visualization.data.group(dt, [1,2],
>>>>>> [{'column': 0, 'aggregation': google.visualization.data.sum, 'type':
>>>>>> 'number'}]);
>>>>>>
>>>>>> Then I tried to tell the API what was data and what was grouping with
>>>>>> the following code, which produced a Column Chart but one in which the
>>>>>> x-axis repeats the year for every country, and the legend has just one
>>>>>> entry with no label:
>>>>>>
>>>>>> var view = new google.visualization.DataView(by_year_country);
>>>>>> view.setColumns([
>>>>>> {sourceColumn:0, type:'string', role:'domain', label:'Year'},
>>>>>> {sourceColumn:1, type:'string', role:'domain', label:'Country'},
>>>>>> {sourceColumn:2, type:'number', role:'data'}
>>>>>> ]);
>>>>>> chart.draw(view, options);
>>>>>>
>>>>>> Can anyone provide any advice on how this is supposed to be
>>>>>> implemented, or the appropriate terminology upon which to seek an answer?
>>>>>>
>>>>>> Many thanks in advance,
>>>>>> Iain
>>>>>>
>>>>>
--
You received this message because you are subscribed to the Google Groups
"Google Visualization API" group.
To view this discussion on the web visit
https://groups.google.com/d/msg/google-visualization-api/-/C-xCnedn4WQJ.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/google-visualization-api?hl=en.
