Merge pull request #9 from ssoudan/0.6.0-rc

0.6.0-rc

.github/workflows/rust.yml

@@ -85,21 +85,21 @@ jobs:
         command: build
         args: --release --all-features --all-targets
 
-  benches:
-    name: Benchmarks
-    runs-on: ubuntu-latest
+  # benches:
+  #   name: Benchmarks
+  #   runs-on: ubuntu-latest
 
-    steps:
-      - name: Install lld
-        run: sudo apt-get update && sudo apt-get install -y lld wget unzip
-      - name: Install latest nightly
-        uses: actions-rs/toolchain@v1
-        with:
-          toolchain: nightly
-          override: true
-      - uses: actions/checkout@v2
-      - name: Bench
-        uses: actions-rs/cargo@v1
-        with:
-          command: bench
-          args: --all --features=unstable
+  #   steps:
+  #     - name: Install lld
+  #       run: sudo apt-get update && sudo apt-get install -y lld wget unzip
+  #     - name: Install latest nightly
+  #       uses: actions-rs/toolchain@v1
+  #       with:
+  #         toolchain: nightly
+  #         override: true
+  #     - uses: actions/checkout@v2
+  #     - name: Bench
+  #       uses: actions-rs/cargo@v1
+  #       with:
+  #         command: bench
+  #         args: --all --features=unstable

.gitignore

@@ -3,3 +3,5 @@
 Cargo.lock
 .vscode
 BUILD.sh
+venv/
+test.ipynb

Cargo.toml

@@ -16,26 +16,30 @@ exclude = [
     "/.cargo/*",
     "/.gitignore",
     "/.devcontainer/*",
+    "/.vscode/*",
+    "/.git/*",
+    "/.idea/*",
+    "/venv/*",
 ]
 
 [features]
 default = []
-unstable = [ "rand", "rand_distr"]
+unstable = ["rand", "rand_distr"]
 
 [[bench]]
 name = "ols"
 harness = false
-required-features = [ "unstable" ]
+required-features = ["unstable"]
 
 [[bench]]
 name = "dickeyfuller"
 harness = false
-required-features = [ "unstable" ]
+required-features = ["unstable"]
 
 [[bench]]
 name = "adf"
 harness = false
-required-features = [ "unstable" ]
+required-features = ["unstable"]
 
 [dependencies]
 nalgebra = "0"

README.md

@@ -8,8 +8,7 @@ Stationarity tests for time-series data in Rust.
 
 At the moment:
 [Dickey-Fuller test](https://en.wikipedia.org/wiki/Dickey–Fuller_test) and 
-[Augmented Dickey-Fuller test](https://en.wikipedia.org/wiki/Augmented_Dickey–Fuller_test) with a 
-constant but no trend.  
+[Augmented Dickey-Fuller test](https://en.wikipedia.org/wiki/Augmented_Dickey–Fuller_test) with no constan no trend, constant or constant and trend.
 
 ## License 
 
@@ -43,16 +42,16 @@ fn main() {
 
     // compute the test statistic
     let regression = Regression::Constant;
-    let report = tools::adf::adf_test(&y, lag, regression).unwrap();
+    let report = tools::adf_test(&y, lag, regression).unwrap();
 
     // critical values for the model with a constant but no trend:
-    let critical_value = distrib::dickeyfuller::get_critical_value(
+    let critical_value: f32 = distrib::dickeyfuller::get_critical_value(
         regression,
         report.size,
         AlphaLevel::OnePercent,
     )
     .unwrap();
-    assert_eq!(report.size, 10);
+    assert_eq!(report.size, 8);
 
     // comparison
     let t_stat = report.test_statistic;

benches/adf.rs

@@ -17,7 +17,7 @@
 use criterion::{criterion_group, criterion_main, BenchmarkId, Criterion};
 use rand::SeedableRng;
 use unit_root::prelude::distrib::Regression;
-use unit_root::prelude::tools::adf::adf_test;
+use unit_root::prelude::tools::adf_test;
 use unit_root::utils::gen_ar_1;
 
 fn adf_benchmark_f32_2(c: &mut Criterion) {

benches/dickeyfuller.rs

@@ -17,7 +17,7 @@
 use criterion::{criterion_group, criterion_main, BenchmarkId, Criterion};
 use rand::SeedableRng;
 use unit_root::prelude::distrib::Regression;
-use unit_root::prelude::tools::dickeyfuller;
+use unit_root::prelude::tools::dickeyfuller_test;
 use unit_root::utils::gen_ar_1;
 
 fn df_benchmark_f32(c: &mut Criterion) {
@@ -30,7 +30,7 @@ fn df_benchmark_f32(c: &mut Criterion) {
         let y = gen_ar_1(&mut rng, *size, mu, delta, sigma);
 
         c.bench_with_input(BenchmarkId::new("df_f32", size), &y, |b, y| {
-            b.iter(|| dickeyfuller::dickeyfuller_test(y, Regression::Constant))
+            b.iter(|| dickeyfuller_test(y, Regression::Constant))
         });
     }
 }
@@ -45,7 +45,7 @@ fn df_benchmark_f64(c: &mut Criterion) {
         let y = gen_ar_1(&mut rng, *size, mu, delta, sigma);
 
         c.bench_with_input(BenchmarkId::new("df_f64", size), &y, |b, y| {
-            b.iter(|| dickeyfuller::dickeyfuller_test(y, Regression::Constant))
+            b.iter(|| dickeyfuller_test(y, Regression::Constant))
         });
     }
 }

examples/adf.rs

@@ -35,7 +35,7 @@ fn main() {
     // compute the test statistic
     let lag = 1;
     let regression = Regression::Constant;
-    let report = tools::adf::adf_test(&y, lag, regression).unwrap();
+    let report = tools::adf_test(&y, lag, regression).unwrap();
 
     // critical values for the model with a constant but no trend:
     let critical_value =

examples/dickeyfuller.rs

@@ -33,7 +33,7 @@ fn main() {
         -0.42968979,
     ]);
 
-    let report = tools::dickeyfuller::dickeyfuller_test(&y, Regression::Constant).unwrap();
+    let report = tools::dickeyfuller_test(&y, Regression::Constant).unwrap();
 
     let critical_value =
         constant_no_trend_critical_value(report.size, AlphaLevel::OnePercent).unwrap();

src/lib.rs

@@ -20,7 +20,7 @@
 //! use unit_root::prelude::distrib::dickeyfuller::get_critical_value;
 //! use unit_root::prelude::distrib::{AlphaLevel, Regression};
 //! use unit_root::prelude::nalgebra::DVector;
-//! use unit_root::prelude::tools::adf::adf_test;
+//! use unit_root::prelude::tools::adf_test;
 //! use unit_root::prelude::*;
 //!
 //! let y = DVector::from_row_slice(&[

src/prelude.rs

@@ -25,17 +25,11 @@ pub use crate::Error;
 
 /// Tools
 pub mod tools {
-    pub use crate::tools::Report;
-
-    /// Dickey-Fuller test
-    pub mod dickeyfuller {
-        pub use crate::tools::dickeyfuller::dickeyfuller_test;
-    }
-
     /// Augmented Dickey-Fuller test
-    pub mod adf {
-        pub use crate::tools::adf::adf_test;
-    }
+    pub use crate::tools::adf::adf_test;
+    /// Dickey-Fuller test
+    pub use crate::tools::dickeyfuller::dickeyfuller_test;
+    pub use crate::tools::Report;
 }
 
 /// Distributions

src/tools/adf.rs

@@ -40,47 +40,112 @@ pub fn adf_test<F: RealField + Scalar + Float>(
     })
 }
 
+/// Comparison with statsmodels.tsa.stattools.adfuller use the following code:
+/// ```python
+/// import numpy as np
+/// import pandas as pd
+/// import statsmodels.tsa.stattools as ts
+/// pd.options.display.float_format = '{:.12g}'.format
+///
+/// def adf_test(timeseries, maxlag=None, regression="c", autolag="AIC"):
+///   print("Results of Dickey-Fuller Test:")
+///   dftest = ts.adfuller(timeseries, maxlag=maxlag, regression=regression, autolag=autolag)
+///   dfoutput = pd.Series(
+///       dftest[0:4],
+///       index=[
+///           "Test Statistic",
+///           "p-value",
+///           "#Lags Used",
+///           "Number of Observations Used",
+///       ],
+///   )
+///   for key, value in dftest[4].items():
+///       dfoutput["Critical Value (%s)" % key] = value
+///   print(dfoutput)
+///
+/// y = [-1.06714348, -1.14700339,  0.79204106, -0.05845247, -0.67476754,
+///      -0.10396661,  1.82059282, -0.51169443,  2.07712365,  1.85668086, 2.56363688]
+///
+/// adf_test(y, maxlag=2, regression='n')
+/// adf_test(y, maxlag=2, regression='c')
+/// adf_test(y, maxlag=2, regression='ct')
+/// ```
+///
+/// Note: this library does not support the `autolag` yet. Tests are using the lag from
+/// statsmodels.
 #[cfg(test)]
 mod tests {
     use approx::assert_relative_eq;
     use nalgebra::DVector;
 
     use crate::distrib::Regression;
-    use crate::prelude::tools::dickeyfuller;
-    use crate::tools::adf::adf_test;
+    use crate::prelude::tools::{adf_test, dickeyfuller_test};
+
+    const Y: [f64; 11] = [
+        -1.06714348,
+        -1.14700339,
+        0.79204106,
+        -0.05845247,
+        -0.67476754,
+        -0.10396661,
+        1.82059282,
+        -0.51169443,
+        2.07712365,
+        1.85668086,
+        2.56363688,
+    ];
+
+    #[test]
+    fn test_t_statistics_n() {
+        let lag = 1;
+        let y = DVector::from_row_slice(&Y[..]);
+
+        let report = adf_test(&y, lag, Regression::NoConstantNoTrend).unwrap();
+        assert_eq!(report.size, 9);
+        assert_relative_eq!(report.test_statistic, -0.417100483298f64, epsilon = 1e-9);
+        // Test Statistic                -0.417100483298
+        // p-value                        0.529851882135
+        // #Lags Used                                  1
+        // Number of Observations Used                 9
+        // Critical Value (1%)                  -2.85894
+        // Critical Value (5%)            -1.96955775034
+        // Critical Value (10%)           -1.58602219479
+    }
 
     #[test]
-    fn test_t_statistics() {
+    fn test_t_statistics_c() {
         let lag = 2;
-        let y = DVector::from_row_slice(&[
-            -1.06714348,
-            -1.14700339,
-            0.79204106,
-            -0.05845247,
-            -0.67476754,
-            -0.10396661,
-            1.82059282,
-            -0.51169443,
-            2.07712365,
-            1.85668086,
-            2.56363688,
-        ]);
+        let y = DVector::from_row_slice(&Y[..]);
 
         let report = adf_test(&y, lag, Regression::Constant).unwrap();
         assert_eq!(report.size, 8);
-        assert_relative_eq!(
-            report.test_statistic,
-            0.48612142266202985f64,
-            epsilon = 1e-3
-        );
-        // statsmodels.tsa.stattools.adfuller(y, maxlag=2, regression='c') gives:
-        // ADF Statistic: 0.486121
-        // p-value: 0.984445
-        // Critical Values:
-        // 1%: -4.665
-        // 5%: -3.367
-        // 10%: -2.803
-        // usedlag= 2
+        assert_relative_eq!(report.test_statistic, 0.486121422662f64, epsilon = 1e-9);
+        // Results of Dickey-Fuller Test:
+        // Test Statistic                0.486121422662
+        // p-value                       0.984445107564
+        // #Lags Used                                 2
+        // Number of Observations Used                8
+        // Critical Value (1%)           -4.66518632812
+        // Critical Value (5%)             -3.367186875
+        // Critical Value (10%)            -2.802960625
+    }
+
+    #[test]
+    fn test_t_statistics_ct() {
+        let lag = 0;
+        let y = DVector::from_row_slice(&Y[..]);
+
+        let report = adf_test(&y, lag, Regression::ConstantAndTrend).unwrap();
+        assert_eq!(report.size, 10);
+        assert_relative_eq!(report.test_statistic, -4.20337098854f64, epsilon = 1e-9);
+        // Results of Dickey-Fuller Test:
+        // Test Statistic                  -4.20337098854
+        // p-value                       0.00442477220907
+        // #Lags Used                                   0
+        // Number of Observations Used                 10
+        // Critical Value (1%)                  -5.282515
+        // Critical Value (5%)                  -3.985264
+        // Critical Value (10%)                  -3.44724
     }
 
     #[test]
@@ -94,7 +159,7 @@ mod tests {
         let regression = Regression::Constant;
 
         let report = adf_test(&y, lag, regression).unwrap();
-        let df_report = dickeyfuller::dickeyfuller_test(&y, regression).unwrap();
+        let df_report = dickeyfuller_test(&y, regression).unwrap();
 
         assert_eq!(report.test_statistic, df_report.test_statistic);
         assert_eq!(report.size, df_report.size);