How can I keep datetime field as one of the axis when I use django-graphos using google API?

[agiliq]

How can I keep datetime field as one of the axis when I use django-graphos using google API?

[AuxProc]

I'm not sure I understand what your current issue is. Could you provide some examples of the data youre giving to django-graphos and what is being returned?

The steps I take to pass datetime to django-graphos using Google looks something like this in my views:

queryset_black_levels = combined_black_levels.order_by('toner_date')[:180]
        if len(queryset_black_levels) > 0:
            haxis_labels = int(len(queryset_black_levels) / 4)
            for item in queryset_black_levels:
                item.toner_date = datetime.date(item.toner_date)
                item.toner_date = item.toner_date.isoformat()

The key is to convert to isoformat. You may also find my chart settings helpful:

chart_black = gchart.LineChart(data_source_black_levels, options={'title': 'Black Levels', 'chartArea': {'width': '75%', 'height': '60%'}, 'legend': { 'position': 'none' }, 'vAxis': { 'format': 'percent', 'viewWindowMode': 'explicit', 'viewWindow': { 'max': 1, 'min': 0}}, 'colors': ['#000000'], 'hAxis': {'format': 'M/d/yy', 'gridlines': {'count': 5}, 'showTextEvery': haxis_labels}}, height=150, width='100%')```

[akshar-raaj]

@recordnotfound Ultimately we need to send this data to frontend and so need to serialize this. So you need to convert your data to something which could be serialized.

You need to override get_data() of existing DataSource and convert datetime field into something which could be serialized.

Assuming you are using a Python list as data, then you need to do:

class MySimpleDataSource(SimpleDataSource):
    def get_data(self):
        data = super(MySimpleDataSource, self).get_data()
        header = data[0]
        data_without_header = data[1:]
        for row in data_without_header:
            # Assuming first column contains datetime
            row[0] = row[0].year
        data_without_header.insert(0, header)
        return data_without_header

And data has

d1 = datetime(2015, 7, 8, 1, 1)
d2 = datetime(2016, 7, 8, 3, 1)

data1 = [
         ['Year', 'Sales', 'Expenses', 'Items Sold', 'Net Profit'],
         [d1, 1000, 400, 100, 600],
         [d2, 1170, 460, 120, 310],
 ]

chart = flot.LineChart(MySimpleDataSource(data=data1))

If you are planning to use queryset with ModelDataSource, then you would create following class

class MyModelDataSource(ModelDataSource):
    def get_data(self):
        data = super(MyModelDataSource, self).get_data()
        header = data[0]
        data_without_header = data[1:]
        for row in data_without_header:
            # Assuming second column contains datetime
            row[1] = row[1].year
        data_without_header.insert(0, header)
        return data_without_header

And you would use this class like:

queryset = Account.objects.all()
chart = flot.LineChart(MyModelDataSource(queryset=queryset, fields=['sales', 'datetime_field','expenses']))

[akshar-raaj]

@recordnotfound It should be doable with the code snippets mentioned in last comment, so closing this ticket. Feel free to open this again if you need further help.

[ndunn219]

In the workaround, @akshar-raaj gets the year, which is serializable, but @recordnotfound asked about a datetime field. I have the same issue using date values. I've tried to replace row[0] = row[0].year with row[0] = row[0].date(), but that returns dates as strings, which JavaScript doesn't recognize as dates. How would you return serializable dates or datetimes?