aeveris/ml-preproc - Change Q7BIMIWQJHBLNB7LYSYAORQPQ2CKRZSOR3VKASIRWBGJWEMICM2AC

Impl HashingVectorizer, add (failing) test

Created by Gz5AN8tiLGjsQNTG24sqRr5jNodcDoAzyZSgZGoaYTaD on July 14, 2021

Q7BIMIWQJHBLNB7LYSYAORQPQ2CKRZSOR3VKASIRWBGJWEMICM2AC

Dependencies

[2] LHETILNNYVQ27JCR5WGRAPDRQBFORK77UDJDMAZ4E47PC7ZPS7RAC

In channels

main

Change contents

Insertion in src/lib.rs at line 5 [2.47]
[2.138]
[2.138]
```
extern crate ndarray_linalg;
```

Replacement in src/lib.rs at line 9 [2.47]

B:BD[2.200] → [2.200:220]

use ndarray::Array;

[2.200]

[2.220]

use ndarray::{Array, Ix2};
use ndarray_linalg::{normalize, Norm, NormalizeAxis};

Replacement in src/lib.rs at line 17 [2.47]
B:BD[2.361] → [2.361:383]
```
    n_samples: usize,
```
[2.361]
[2.383]
```
    n_features: u32,
```

Replacement in src/lib.rs at line 21 [2.47]

B:BD[2.411] → [2.411:482]

    pub fn new(n_samples: usize, n_min: usize, n_max: usize) -> Self {

[2.411]

[2.482]

    pub fn new(n_features: u32, n_min: usize, n_max: usize) -> Self {

Replacement in src/lib.rs at line 23 [2.47]
B:BD[2.497] → [2.497:520]
```
            n_samples,
```
[2.497]
[2.520]
```
            n_features,
```

Replacement in src/lib.rs at line 29 [2.47]

B:BD[2.575] → [2.575:625]

    pub fn fit_transform(&self, input: &[&str]) {

[2.575]

[2.625]

    pub fn fit_transform(&self, input: &[&str]) -> Array<f64, Ix2> {

Replacement in src/lib.rs at line 31 [2.47]

B:BD[2.660] → [2.660:758]

        let mut target_arr: Array<f64, _> = ndarray::Array::zeros((input.len(), self.n_samples));

[2.660]

[2.758]


        let mut target: Array<f64, _> = Array::zeros((input.len(), self.n_features as usize));

Replacement in src/lib.rs at line 35 [2.47]

B:BD[2.811] → [2.811:872]

            for mut ng in NGramIter::new(&normalize(s), 5) {

[2.811]

[2.872]

            let mut v = target.row_mut(idx);
            assert_eq!(v.shape(), &[16]);
            for ng in NGramIter::new(&normalize_str(s), 5) {

Insertion in src/lib.rs at line 39 [2.47]

[2.942]

                let f_idx = h % self.n_features;
                // If the highest bit is 0, we want switch to be a 1
                let switch = (!h & 0x80_00_00_00) >> 31;
                v[f_idx as usize] += (switch * 2) as f64 - 1.0;

Insertion in src/lib.rs at line 45 [2.47]

[2.966]

        println!("{:?}", target);
        let target = normalize(target, NormalizeAxis::Row).0;
        println!("{:?}", target);
        target

Replacement in src/lib.rs at line 52 [2.47]
B:BD[2.975] → [2.975:1013]
```
fn normalize(input: &str) -> String {
```
[2.975]
[2.1013]
```
fn normalize_str(input: &str) -> String {
```

Replacement in src/lib.rs at line 127 [2.47]

B:BD[2.2869] → [2.2869:2895]

    use crate::NGramIter;

[2.2869]

[2.2895]

    use ndarray::{array, Array, Ix2};
    use crate::{HashingVectorizer, NGramIter};

Insertion in src/lib.rs at line 140 [2.47]

[2.3250]

    }
    #[test]
    fn vect_test() {
        let v = HashingVectorizer::new(16, 5, 5);
        let input = [
            "This is the first document.",
            "This document is the second document.",
            "And this is the third one.",
            "Is this the first document?",
        ];
        let a = v.fit_transform(&input);
        //assert_eq!(val, Vec::new());
        let t: Array<f64, _> = Array::zeros((input.len(), 16));
        assert_eq!(t, a);

Replacement in Cargo.toml at line 10 [2.3294]

B:BD[2.3490] → [2.3490:3535]

ndarray = "0.15"
unicode-normalization ="0.1"

[2.3490]

ndarray = "0.14"
ndarray-linalg = { version = "0.13" }
unicode-normalization = "0.1"

Insertion in Cargo.lock at line 6 [2.3571]

[2.3688]

name = "approx"
version = "0.4.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3f2a05fd1bd10b2527e20a2cd32d8873d115b8b39fe219ee25f42a8aca6ba278"
dependencies = [
 "num-traits",
]
[[package]]

Insertion in Cargo.lock at line 19 [2.3571]

[2.3866]


[[package]]
name = "blas-src"
version = "0.6.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "a25f136d1bcc6fc28330077511a0646aaf75bc894e5532b1a33a93d814272328"
[[package]]
name = "cauchy"
version = "0.3.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3c06b9e1a9ea243aa7be8cf51e025562334a52674508cc02d8b54a99d4986a13"
dependencies = [
 "num-complex",
 "num-traits",
 "rand",
 "serde",
]
[[package]]
name = "cblas-sys"
version = "0.1.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b6feecd82cce51b0204cf063f0041d69f24ce83f680d87514b004248e7b0fa65"
dependencies = [
 "libc",
]
[[package]]
name = "cfg-if"
version = "1.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "baf1de4339761588bc0619e3cbc0120ee582ebb74b53b4efbf79117bd2da40fd"
[[package]]
name = "getrandom"
version = "0.1.16"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8fc3cb4d91f53b50155bdcfd23f6a4c39ae1969c2ae85982b135750cccaf5fce"
dependencies = [
 "cfg-if",
 "libc",
 "wasi",
]

Insertion in Cargo.lock at line 65 [2.3571]

[2.3879]

name = "lapack"
version = "0.17.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "041b3cb380f92a53e317ea046d0e715d2302dae6d2f426bdf3ccf27a36ef40e9"
dependencies = [
 "lapack-sys",
 "libc",
 "num-complex",
]
[[package]]
name = "lapack-sys"
version = "0.12.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "c1d3a8a9f07310243de6c6226f039f14bce8d2f4c96b5d30ddbcfa31eb4e94ad"
dependencies = [
 "libc",
]
[[package]]
name = "lax"
version = "0.1.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "592fa21cdc5f1be099c58a64d2e51a9c02632f4c2b7fb4ee8c2994d4950a4ff7"
dependencies = [
 "cauchy",
 "lapack",
 "num-traits",
 "thiserror",
]
[[package]]
name = "libc"
version = "0.2.98"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "320cfe77175da3a483efed4bc0adc1968ca050b098ce4f2f1c13a56626128790"
[[package]]

Replacement in Cargo.lock at line 104 [2.3571]
B:BD[2.3903] → [2.3903:3921]
```
version = "0.3.1"
```
[2.3903]
[2.3921]
```
version = "0.2.4"
```

Replacement in Cargo.lock at line 106 [2.3571]

B:BD[2.3986] → [2.3986:4064]

checksum = "5a8a15b776d9dfaecd44b03c5828c2199cddff5247215858aac14624f8d6b741"

[2.3986]

[2.4064]

checksum = "916806ba0031cd542105d916a97c8572e1fa6dd79c9c51e7eb43a09ec2dd84c1"

Insertion in Cargo.lock at line 117 [2.3571]
[2.4190]
[2.4190]
```
 "ndarray-linalg",
```
Replacement in Cargo.lock at line 129 [2.3571]
B:BD[2.4439] → [2.4439:4458]
```
version = "0.15.3"
```
[2.4439]
[2.4458]
```
version = "0.14.0"
```

Replacement in Cargo.lock at line 131 [2.3571]

B:BD[2.4523] → [2.4523:4601]

checksum = "08e854964160a323e65baa19a0b1a027f76d590faba01f05c0cbc3187221a8c9"

[2.4523]

[2.4601]

checksum = "6c0d5c9540a691d153064dc47a4db2504587a75eae07bf1d73f7a596ebc73c04"

Insertion in Cargo.lock at line 133 [2.3571]
[2.4618]
[2.4618]
```
 "approx",
 "blas-src",
 "cblas-sys",
```

Insertion in Cargo.lock at line 144 [2.3571]

[2.4714]

name = "ndarray-linalg"
version = "0.13.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "dfb88fe83ac017c544b6042d172e62a4d1e9b293e22caf46ae305755cc533a7a"
dependencies = [
 "cauchy",
 "lax",
 "ndarray",
 "num-complex",
 "num-traits",
 "rand",
 "thiserror",
]
[[package]]

Replacement in Cargo.lock at line 160 [2.3571]
B:BD[2.4735] → [2.4735:4753]
```
version = "0.4.0"
```
[2.4735]
[2.4753]
```
version = "0.3.1"
```

Replacement in Cargo.lock at line 162 [2.3571]

B:BD[2.4818] → [2.4818:4896]

checksum = "26873667bbbb7c5182d4a37c1add32cdf09f841af72da53318fdb81543c15085"

[2.4818]

[2.4896]

checksum = "747d632c0c558b87dbabbe6a82f3b4ae03720d0646ac5b7b4dae89394be5f2c5"

Insertion in Cargo.lock at line 165 [2.3571]
[2.4928]
[2.4928]
```
 "rand",
 "serde",
```

Insertion in Cargo.lock at line 186 [2.3571]

[2.5396]

]
[[package]]
name = "ppv-lite86"
version = "0.2.10"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ac74c624d6b2d21f425f752262f42188365d7b8ff1aff74c82e45136510a4857"
[[package]]
name = "proc-macro2"
version = "1.0.27"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f0d8caf72986c1a598726adc988bb5984792ef84f5ee5aa50209145ee8077038"
dependencies = [
 "unicode-xid",
]
[[package]]
name = "quote"
version = "1.0.9"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "c3d0b9745dc2debf507c8422de05d7226cc1f0644216dfdfead988f9b1ab32a7"
dependencies = [
 "proc-macro2",
]
[[package]]
name = "rand"
version = "0.7.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "6a6b1679d49b24bbfe0c803429aa1874472f50d9b363131f0e89fc356b544d03"
dependencies = [
 "getrandom",
 "libc",
 "rand_chacha",
 "rand_core",
 "rand_hc",
]
[[package]]
name = "rand_chacha"
version = "0.2.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f4c8ed856279c9737206bf725bf36935d8666ead7aa69b52be55af369d193402"
dependencies = [
 "ppv-lite86",
 "rand_core",

Insertion in Cargo.lock at line 236 [2.3571]

[2.5411]

name = "rand_core"
version = "0.5.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "90bde5296fc891b0cef12a6d03ddccc162ce7b2aff54160af9338f8d40df6d19"
dependencies = [
 "getrandom",
]
[[package]]
name = "rand_hc"
version = "0.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ca3129af7b92a17112d59ad498c6f81eaf463253766b90396d39ea7a39d6613c"
dependencies = [
 "rand_core",
]
[[package]]

Insertion in Cargo.lock at line 258 [2.3571]

[2.5592]


[[package]]
name = "serde"
version = "1.0.126"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ec7505abeacaec74ae4778d9d9328fe5a5d04253220a85c4ee022239fc996d03"

Insertion in Cargo.lock at line 266 [2.3571]

[2.5605]

name = "syn"
version = "1.0.73"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f71489ff30030d2ae598524f61326b902466f72a0fb1a8564c001cc63425bcc7"
dependencies = [
 "proc-macro2",
 "quote",
 "unicode-xid",
]
[[package]]
name = "thiserror"
version = "1.0.26"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "93119e4feac1cbe6c798c34d3a53ea0026b0b1de6a120deef895137c0529bfe2"
dependencies = [
 "thiserror-impl",
]
[[package]]
name = "thiserror-impl"
version = "1.0.26"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "060d69a0afe7796bf42e9e2ff91f5ee691fb15c53d38b4b62a9a53eb23164745"
dependencies = [
 "proc-macro2",
 "quote",
 "syn",
]
[[package]]

Insertion in Cargo.lock at line 319 [2.3571]

[2.6256]


[[package]]
name = "unicode-xid"
version = "0.2.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8ccb82d61f80a663efe1f787a51b16b5a51e3314d6ac365b08639f52387b33f3"
[[package]]
name = "wasi"
version = "0.9.0+wasi-snapshot-preview1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "cccddf32554fecc6acb585f82a32a72e28b48f8c4c1883ddfeeeaa96f7d8e519"