Error creating string ND array

[19csi]

Hi,

I'm trying to create an NDarray to store DateTime values from a normal C# array. When I try to create the NDarray I get the following error: Can not convert type of given object to dtype: System.String[].

            string[] times = new string[50];

            DateTime now = DateTime.Now;

            for (int i = 0; i < 50; i++)            
                times[i] = now.AddSeconds(i).ToString();            

            // Raises Exception
            NDarray<string> nTimes = np.array<string>(times);

When I try to make a NDarray<float> it works fine, but for NDarray<DateTime> and NDarray<string> this error occurs. Is there something I am doing wrong here or are these types not supported? Thanks for the help!

[henon]

Yeah, DateTime and String can't simply be marshalled into Python like Integers or Floats, so I left support for them out for now. I'll look into it. Hope it is not too complicated.

[19csi]

Gotcha, that's what I was thinking but wanted to make sure. Thanks for the quick response.

[henon]

with a little modification an array of strings is already working:

        [TestMethod]
        public void StringArray()
        {
            //>>> a = numpy.array(['apples', 'foobar', 'cowboy'])
            //>>> a[2] = 'bananas'
            //>>> a
            //array(['apples', 'foobar', 'banana'], 
            //      dtype = '|S6')
            var a = np.array(new string[]{"apples", "foobar", "cowboy"});
            Assert.AreEqual("array(['apples', 'foobar', 'cowboy'], dtype='<U6')", a.repr);
            // todo: a[2]="banana";
            a.self.SetItem(new PyInt(2), new PyString("banana"));
            Assert.AreEqual("array(['apples', 'foobar', 'banana'], dtype='<U6')", a.repr);

            //>>> a = numpy.array(['apples', 'foobar', 'cowboy'], dtype = object)
            //>>> a
            //array([apples, foobar, cowboy], dtype = object)
            //>>> a[2] = 'bananas'
            //>>> a
            //array([apples, foobar, bananas], dtype = object)
            a = np.array(new string[] { "apples", "foobar", "cowboy" }, dtype:np.object_);
            Assert.AreEqual("array(['apples', 'foobar', 'cowboy'], dtype=object)", a.repr);
            // todo: a[2]="banana";
            a.self.SetItem(new PyInt(2), new PyString("banana"));
            Assert.AreEqual("array(['apples', 'foobar', 'banana'], dtype=object)", a.repr);
        }

is that enough for your use case?

[henon]

in any case, I released the change as v21 to nuget

[19csi]

Awesome, that works for me! That was fast too 👍

[awa5114]

@henon I wouldn't suppose any workarounds exist to getting something like this to work?

NDarray<DateTime> dateTimes = np.array( new DateTime[] {
    new DateTime(2000, 1, 15),
    new DateTime(2000, 6, 15),
    new DateTime(2001, 1, 15),
    new DateTime(2001, 6, 15)
});

[henon]

@amine-aboufirass I created a separate issue for your request: #119