v0.2.4

PharmCat · Feb 8, 2021 · b420fc7 · b420fc7
1 parent 70ed69d
commit b420fc7
Show file tree

Hide file tree

Showing 17 changed files with 108 additions and 409 deletions.
diff --git a/Project.toml b/Project.toml
@@ -1,7 +1,7 @@
 name = "Metida"
 uuid = "a1dec852-9fe5-11e9-361f-8d9fde67cfa2"
 authors = ["Vladimir Arnautov <[email protected]>"]
-version = "0.2.3"
+version = "0.2.4"
 
 [deps]
 CategoricalArrays = "324d7699-5711-5eae-9e2f-1d82baa6b597"

diff --git a/change.log b/change.log
@@ -1,6 +1,6 @@
-- 0.2.4
+v0.2.4
   * minor optimizations
-  * docs
+  * docs remake
   * split test and validation
   * new test datasets
 

diff --git a/docs/make.jl b/docs/make.jl
@@ -8,11 +8,13 @@ makedocs(
     doctest = true,
     pages = [
         "Home" => "index.md",
-        "Details" => "details.md",
+        "User guide" => [
+        "First step" => "instanduse.md",
         "Examples" => "examples.md",
-        "Options" => "options.md",
         "NLopt" => "nlopt.md",
         "CUDA" => "cuda.md",
+        ],
+        "Details" => "details.md",
         "Benchmark" => "bench.md",
         "Validation" => "validation.md",
         "API" => "api.md",

diff --git a/docs/src/api.md b/docs/src/api.md
@@ -1,4 +1,4 @@
-## API
+# API
 
 ### Metida.LMM
 ```@docs

diff --git a/docs/src/bench.md b/docs/src/bench.md
@@ -1 +1 @@
-## Benchmark
+# Benchmark
diff --git a/docs/src/cuda.md b/docs/src/cuda.md
@@ -1,4 +1,4 @@
-## CUDA
+### CUDA
 
 Use CuBLAS & CuSOLVER for REML calculation. Optimization with NLopt.jl.
 

diff --git a/docs/src/details.md b/docs/src/details.md
@@ -1,6 +1,6 @@
-## Details
+# Details
 
-The solution to the mixed model equations is a maximum likelihood estimate when the distribution of the errors is normal. Maximum likelihood estimates are based on the probability model for the observed responses. In the probability model the distribution of the responses is expressed as a function of one or more parameters. PROC MIXED in SAS used restricted maximum likelihood (REML) approach by default. REML equation can be described with following (Henderson,  1959;Laird et.al. 1982; Jennrich 1986; Lindstrom & Bates, 1988; Gurka et.al 2006). 
+The solution to the mixed model equations is a maximum likelihood estimate when the distribution of the errors is normal. Maximum likelihood estimates are based on the probability model for the observed responses. In the probability model the distribution of the responses is expressed as a function of one or more parameters. PROC MIXED in SAS used restricted maximum likelihood (REML) approach by default. REML equation can be described with following (Henderson,  1959;Laird et.al. 1982; Jennrich 1986; Lindstrom & Bates, 1988; Gurka et.al 2006).
 
 Metida.jl using optimization with Optim.jl package (Newton's Method) by default.  Because variance have only positive values and ρ is limited as -1 ≤ ρ ≤ 1 in Metida.jl "link" function is used. Exponential values is optimizing in variance part and ρ is linked with sigmoid function.
 All steps perform with differentiable functions with forward automatic differentiation using ForwardDiff.jl package. Also [MetidaNLopt.jl](https://github.com/PharmCat/MetidaNLopt.jl) and [MetidaCu.jl](https://github.com/PharmCat/MetidaCu.jl) available for optimization with NLopt.jl and solving on CUDA GPU. Sweep algorithm using for variance-covariance matrix inversing in REML calculation.