Update docs

.github/workflows/CI.yml

@@ -24,7 +24,7 @@ jobs:
       matrix:
         version:
           - '1.11'
-          - '1.6'
+          - '1'
           - 'pre'
         os:
           - ubuntu-latest

Dockerfile

@@ -8,7 +8,7 @@ WORKDIR /tmp
 COPY . /tmp
 
 # Preinstall the dependencies in the global environment
-RUN julia --project=. -e "using Pkg; Pkg.instantiate(); Pkg.precompile()"
+RUN julia --project=. -e "using Pkg; Pkg.instantiate(); Pkg.precompile(); Pkg.test();"
 
 # Set the default working directory for the container
 WORKDIR /workdir

Project.toml

@@ -1,13 +1,10 @@
 name = "DFA"
-uuid = "68b8aea9-720c-4092-afc2-f92dd305eeec"
+uuid = "2556a7c9-976d-46bb-b923-66b4a6f0a5bd"
 authors = ["afternone", "Alberto Barradas <[email protected]>"]
 version = "v1.0.0-DEV"
 
-[deps]
-Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
 
 [compat]
-Statistics = "1.11.1"
 julia = "1"
 
 [extras]

README.md

@@ -43,7 +43,7 @@ scales, fluc = dfa(x, boxmax=1000, boxmin=4, boxratio=1.2, overlap=0.5)
 ```
 To estimates the scaling exponent:
 ```
-intercept, α = polyfit(log10(scales), log10(fluc))  # α is scaling exponent
+intercept, α = polyfit(log10.(scales), log10.(fluc))  # α is scaling exponent
 ```
 To plot F(n):
 ```
@@ -53,8 +53,8 @@ scatter(scales, fluc, "o")
 ```
 To plot F(n) with fitted line:
 ```
-log_scales = log10(scales)
-plot(log_scales, log10(fluc), "o", log_scales, α.*log_scales.+intercept)
+log_scales = log10.(scales)
+plot(log_scales, log10.(fluc), "o", log_scales, α.*log_scales.+intercept)
 ```
 
 ## References

docs/Project.toml

@@ -1,4 +1,7 @@
 [deps]
 DFA = "2556a7c9-976d-46bb-b923-66b4a6f0a5bd"
 Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
+DocumenterTools = "35a29f4d-8980-5a13-9543-d66fff28ecb8"
+HTTP = "cd3eb016-35fb-5094-929b-558a96fad6f3"
+LiveServer = "16fef848-5104-11e9-1b77-fb7a48bbb589"
 Plots = "91a5bcdd-55d7-5caf-9e0b-520d859cae80"

docs/src/index.md

@@ -2,6 +2,10 @@
 CurrentModule = DFA
 ```
 
+<!-- '''@setup
+using Random: seed!
+seed!(137)
+''' -->
 # DFA
 
 Documentation for [DFA](https://github.com/abcsds/DFA.jl).
@@ -20,29 +24,35 @@ pkg> add https://github.com/abcsds/DFA.jl
 
 ## Usage
 
-### Basic Example
+### Basic usage
 
 We'll perform a DFA and estimate the scaling exponent for a random time series.
 
 ```julia
-using DFA
-
 x = rand(10000)
+
 scales, fluc = dfa(x)
-```
+log_scales = log10.(scales)
+log_fluc = log10.(fluc)
+intercept, α = polyfit(log_scales, log_fluc)
+α
 
-### Advanced Example
+# output
 
-To perform a DFA on `x` with `boxmax=1000`, `boxmin=4`, `boxratio=1.2`, `overlap=0.5`:
+0.49706362711274654
 
-```julia
-scales, fluc = dfa(x, boxmax=1000, boxmin=4, boxratio=1.2, overlap=0.5)
 ```
 
+
 To estimate the scaling exponent:
 
 ```julia
 intercept, α = polyfit(log10(scales), log10(fluc))  # α is the scaling exponent
+α
+
+# output
+
+0.49706362711274654
 ```
 
 To plot F(n):
@@ -61,6 +71,37 @@ plot(log_scales, log10(fluc), seriestype=:scatter, label="DFA")
 plot!(log_scales, α .* log_scales .+ intercept, label="Fit")
 ```
 
+
+### Advanced Example
+
+To perform a DFA on heart rate variability data `x`, two different box sizes are commonly used for extracting two coefficients, α1 and α2. The first box size is 4 ≤ n ≤ 16, and the second box size is 16 ≤ n ≤ 64. The scaling exponent α1 is used to quantify the short-term correlation properties of the time series, while α2 is used to quantify the long-term correlation properties of the time series.
+
+```julia
+using HTTP
+
+url = "https://physionet.org/files/rr-interval-healthy-subjects/1.0.0/000.txt"
+hrv = parse.(Float64, filter!(e->e!="", split(String(HTTP.get(url).body), r"[^\d.]")))
+x = hrv
+
+scales, fluc = dfa(x, boxmax=64, boxmin=4, boxratio=2, overlap=0.0 )
+log_scales = log10.(scales)
+log_fluc = log10.(fluc)
+intercept, α1 = polyfit(log_scales, log_fluc)
+
+scales, fluc = dfa(x, boxmax=16, boxmin=4, boxratio=2, overlap=0.0 )
+log_scales = log10.(scales)
+log_fluc = log10.(fluc)
+intercept, α2 = polyfit(log_scales, log_fluc)
+
+println("Physionet rr-interval-healthy-subjects/1.0.0/000 α1: $α1")
+println("Physionet rr-interval-healthy-subjects/1.0.0/000 α2: $α2")
+(α1, α2)
+
+# output
+
+(1.096221478218021, 1.0773259128065056)
+```
+
 ## Functions
 
 ### `dfa`
@@ -105,5 +146,5 @@ e220.
 ```
 
 ```@autodocs
-Modules = [DFA Plots]
+Modules = [DFA]
 ```

test/Project.toml

@@ -7,5 +7,4 @@ Plots = "91a5bcdd-55d7-5caf-9e0b-520d859cae80"
 Profile = "9abbd945-dff8-562f-b5e8-e1ebf5ef1b79"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 SignalAnalysis = "df1fea92-c066-49dd-8b36-eace3378ea47"
-Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"

test/runtests.jl

@@ -21,7 +21,7 @@ import HTTP
         intercept, α = polyfit(log10.(scales), log10.(fluc))
         @test intercept ≈ -0.645847540086 atol=1e-12
         @test α ≈ 0.509373252211 atol=1e-12
-        println("Simple Normal distribution: $intercept, $α")
+        # println("Simple Normal distribution: $intercept, $α")
     end
 
     # Test wite noise
@@ -35,7 +35,7 @@ import HTTP
         intercept, α = polyfit(log10.(scales), log10.(fluc))
         @test intercept ≈ -1.1746914232962604 atol=1e-12
         @test α ≈ 0.5237515930298794 atol=1e-12
-        println("Uniform distribution: $intercept, $α")
+        # println("Uniform distribution: $intercept, $α")
     end
 
     # Test random walk / brownian motion
@@ -47,7 +47,7 @@ import HTTP
         intercept, α = polyfit(log10.(scales), log10.(fluc))
         @test intercept ≈ -1.4091958750176548 atol=1e-12
         @test α ≈ 1.5477976383996574 atol=1e-12
-        println("Brownian motion: $intercept, $α")
+        # println("Brownian motion: $intercept, $α")
     end
 
     # Test pink noise
@@ -59,7 +59,7 @@ import HTTP
         intercept, α = polyfit(log10.(scales), log10.(fluc))
         @test intercept ≈ -1.2717558431479488 atol=1e-12
         @test α ≈ 1.0161132560075559 atol=1e-12
-        println("Pink noise: $intercept, $α")
+        # println("Pink noise: $intercept, $α")
     end
 
     # Test wite noise
@@ -72,7 +72,7 @@ import HTTP
             intercept, α = polyfit(log10.(scales), log10.(fluc))
             @test intercept ≈ 0.10576185464214086 atol=1e-12
             @test α ≈ 1.5734083037634519 atol=1e-12
-            println("Under scale (300) and translate(sin(2πx/100)): $intercept, $α")
+            # println("Under scale (300) and translate(sin(2πx/100)): $intercept, $α")
         end
 
         @testset "DFA.transforms.translate" begin
@@ -83,7 +83,7 @@ import HTTP
             intercept, α = polyfit(log10.(scales), log10.(fluc))
             @test intercept ≈ -2.6458475400862542 atol=1e-12
             @test α ≈ 0.5093732522114474 atol=1e-12
-            println("Under translate(300): $intercept, $α")
+            # println("Under translate(300): $intercept, $α")
         end
     end
 
@@ -160,7 +160,7 @@ import HTTP
         intercept, α = polyfit(log_scales, log_fluc)
         @test intercept ≈ 0.6476733308961827
         @test α ≈ 1.0377546755013551 atol=1e-12
-        println("Physionet rr-interval-healthy-subjects/1.0.0/000: $intercept, $α")
+        # println("Physionet rr-interval-healthy-subjects/1.0.0/000: $intercept, $α")
 
         Plots.plot(log_scales, log_fluc, label="DFA", xlabel="Scale Log10", ylabel="Fluctuation Log10", seriestype=:scatter)
         Plots.plot!(log_scales, intercept .+ α .* log_scales, label="DFA Fit")
@@ -195,8 +195,8 @@ import HTTP
         Plots.plot!(log_scales, intercept .+ α1 .* log_scales, label="DFA Fit α1")
         Plots.plot!(log_scales, intercept .+ α2 .* log_scales, label="DFA Fit α2")
         Plots.savefig("DFA.png")
-        println("Physionet rr-interval-healthy-subjects/1.0.0/000 α1: $α1")
-        println("Physionet rr-interval-healthy-subjects/1.0.0/000 α2: $α2")
+        # println("Physionet rr-interval-healthy-subjects/1.0.0/000 α1: $α1")
+        # println("Physionet rr-interval-healthy-subjects/1.0.0/000 α2: $α2")
 
     end
 end