arrayfire_fork 3.8.1

ArrayFire is a high performance software library for parallel computing with an easy-to-use API. Its array based function set makes parallel programming simple. ArrayFire's multiple backends (CUDA, OpenCL and native CPU) make it platform independent and highly portable. A few lines of code in ArrayFire can replace dozens of lines of parallel computing code, saving you valuable time and lowering development costs. This crate provides Rust bindings for ArrayFire library.
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

# Serialization & Deserialization of ArrayFire Objects

To save [Array][1] contents, shape in JSON format, it just takes couple of lines of code as shown below:
```rust,noplaypen
{{#include ../../src/core/array.rs:array_json_serde_snippet}}
```
Saving [Array][1] in different formats is as simple as changing the object qualifier of methods `serialize` and `deserialize`. For example, if user wants to store [Array][1] in `bincode` format instead of JSON, the above code only needs to be change in couple of lines.
```rust,noplaypen
{{#include ../../src/core/array.rs:array_bincode_serde_snippet}}
```

In similar fashion, we can serialize and deserialize [Dim4][2], [RandomEngine][3], [Seq][4] and other Enums. Examples of [Dim4][2], [RandomEngine][3] and [Seq][4] are given below.

```rust,noplaypen
{{#include ../../src/core/dim4.rs:dim4_json_serde_snippet}}
```

```rust,noplaypen
{{#include ../../src/core/random.rs:rng_bincode_serde_snippet}}
```

```rust,noplaypen
{{#include ../../src/core/seq.rs:seq_json_serde_snippet}}
```

[1]: http://arrayfire.org/arrayfire-rust/arrayfire/struct.Array.html
[2]: http://arrayfire.org/arrayfire-rust/arrayfire/struct.Dim4.html
[3]: http://arrayfire.org/arrayfire-rust/arrayfire/struct.RandomEngine.html
[4]: http://arrayfire.org/arrayfire-rust/arrayfire/struct.Seq.html