Most efficient Django model to store numerical data

Question

What is the best way to store numerical data as a model in Django? I did look at What is the most efficent way to store a list in the Django models?, but I'd be curious to hear an answer tailored to numpy arrays as well.

Answer 1

One way would be to convert the numpy array to a string and store it in a textfield

base64.encodestring(nparray)

Another way would be to dump an array into a file and storing the path of the text file in the database

nparray.dump(file)

If you want to store the data in a structured manner in Django, you will need to create models to do that.

You could use 2 Models:

class Element(models.Model):
    Value = models.FloatField()
    Array = models.ForeignKey(Array)

class Array(models.Model):
    #Not required, just for illustration, use the id models instead
    Name = models.CharField('Name', max_length=100)

    Parent = models.ForeignKey(self, blank=True, null=True)

You store the values in the Element model and create the structure using the Array model.

Say you have a 2D array you can store it this way [Array1, Array2, Array3] Array1 = [1,2,3] Array2 = [4,5,6] Array3 = [7,8,9]

Array('ParnetArray')

Array('Array1','ParentArray'),Array('Array2','ParentArray'),Array('Array3','ParentArray')

Element(1,'Array1'),Element(2,'Array1'),Element(3,'Array1'),Element(4',Array2'),Element(5,'Array2')...........