1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
use crate::internal::*;
use ndarray::*;

use super::MultiBroadcastTo;

#[derive(Debug, Clone, new, Default, Hash)]
pub struct Tile {
    pub multipliers: TVec<TDim>,
}

impl Tile {
    fn eval_t<T: Datum>(data: &TValue, multipliers: &[usize]) -> TractResult<TValue> {
        let view = unsafe { data.to_array_view_unchecked::<T>() };
        let output_shape: TVec<usize> =
            view.shape().iter().zip(multipliers.iter()).map(|(&d, &m)| d * m).collect();
        let output = ndarray::ArrayD::from_shape_fn(&*output_shape, |coords| {
            let coords: TVec<usize> =
                coords.slice().iter().zip(data.shape().iter()).map(|(&x, &d)| x % d).collect();
            view[&*coords].clone()
        });
        let mut output = output.into_tensor();
        unsafe {
            output.set_datum_type(data.datum_type());
        }

        Ok(output.into_tvalue())
    }
}

impl Op for Tile {
    fn name(&self) -> Cow<str> {
        "Tile".into()
    }

    fn info(&self) -> TractResult<Vec<String>> {
        Ok(vec![format!("multipliers: {:?}", self.multipliers)])
    }

    op_as_typed_op!();
}

impl EvalOp for Tile {
    fn is_stateless(&self) -> bool {
        true
    }

    fn eval_with_session(
        &self,
        session: &SessionState,
        inputs: TVec<TValue>,
    ) -> TractResult<TVec<TValue>> {
        let multipliers: TVec<usize> = self
            .multipliers
            .iter()
            .map(|m| m.eval(&session.resolved_symbols).to_usize())
            .collect::<TractResult<_>>()?;
        let result = dispatch_datum_by_size!(Self::eval_t(inputs[0].datum_type())(
            &inputs[0],
            &multipliers
        ))?;
        Ok(tvec!(result))
    }

}

impl TypedOp for Tile {
    as_op!();

    fn concretize_dims(
        &self,
        _source: &TypedModel,
        node: &TypedNode,
        target: &mut TypedModel,
        mapping: &HashMap<OutletId, OutletId>,
        values: &SymbolValues,
    ) -> TractResult<TVec<OutletId>> {
        let multipliers = self.multipliers.iter().map(|m| m.eval(values)).collect();
        target.wire_node(&node.name, Self { multipliers }, &[mapping[&node.inputs[0]]])
    }

    fn declutter(
        &self,
        model: &TypedModel,
        node: &TypedNode,
    ) -> TractResult<Option<TypedModelPatch>> {
        let input_fact = model.outlet_fact(node.inputs[0])?;
        if input_fact
            .shape
            .iter()
            .zip(self.multipliers.iter())
            .all(|(i, m)| i.is_one() || m.is_one())
        {
            let output_fact = self.output_facts(&[input_fact])?.remove(0);
            TypedModelPatch::replace_single_op(
                model,
                node,
                &node.inputs,
                MultiBroadcastTo { shape: output_fact.shape },
            )
            .map(Some)
        } else {
            Ok(None)
        }
    }

    fn output_facts(&self, inputs: &[&TypedFact]) -> TractResult<TVec<TypedFact>> {
        let shape = inputs[0]
            .shape
            .iter()
            .zip(self.multipliers.iter())
            .map(|(a, b)| a.clone() * b)
            .collect::<TVec<_>>();
        Ok(tvec!(inputs[0].datum_type.fact(shape)))
    }
}