How to iterate the dataset with next()?

Ray Libraries (Data, Train, Tune, Serve) Ray Data
1

ds = ray.data.from_items([
{“food”: “spam”, “price”: 9.34},
{“food”: “ham”, “price”: 5.37},
{“food”: “eggs”, “price”: 0.94}
])

Now ray supports for loop in batches as below:

for batch in ds.iter_batches().

Suppose I want to iterate the batches manually with next() , is it possbile ?

2

Yes, it should be possible. Something like:

iterator = iter(ds.iter_batches(...))
next(iterator)

should work.

4

No, it doesn’t work. I have tried it .

5

This worked for me:

import ray
>>> ds = ray.data.from_items([
... {"food": "spam", "price": 9.34},
... {"food": "ham", "price": 5.37},
... {"food": "eggs", "price": 0.94}
... ])
>>> iterator = iter(ds.iter_batches(batch_size=1))
>>> next(iterator)
{'food': array(['spam'], dtype=object), 'price': array([9.34])}
6

My falut. I have tried

 iterator = ds.iterator()
 print(next(iterator))

it didn’t work. But your code works well. Thanks a lot.