Merge pull request #325 from nfdi4plants/performanceReport

Performance report

README.md

@@ -1,19 +1,16 @@
 # ARCtrl
 
-> **ARCtrl** the easy way to read, manipulate and write ARCs in __.NET__ 
-and __JavaScript__! ❤️
+> **ARCtrl** the easy way to read, manipulate and write ARCs in __.NET__, __JavaScript__ and __Python__! ❤️
 
 | Version | Downloads |
 | :--------|-----------:|
 |<a href="https://www.nuget.org/packages/ARCtrl/"><img alt="Nuget" src="https://img.shields.io/nuget/vpre/ARCtrl?logo=nuget&color=%234fb3d9"></a>|<a href="https://www.nuget.org/packages/ARCtrl/"><img alt="Nuget" src="https://img.shields.io/nuget/dt/ARCtrl?color=%234FB3D9"></a>|
 |<a href="https://www.npmjs.com/package/@nfdi4plants/arctrl"><img alt="NPM" src="https://img.shields.io/npm/v/%40nfdi4plants/arctrl?logo=npm&color=%234fb3d9"></a>|<a href="https://www.npmjs.com/package/@nfdi4plants/arctrl"><img alt="NPM" src="https://img.shields.io/npm/dt/%40nfdi4plants%2Farctrl?color=%234fb3d9"></a>|
-
+|<a href="https://pypi.org/project/ARCtrl/"><img alt="PyPI" src="https://img.shields.io/pypi/v/arctrl?logo=pypi&color=%234fb3d9"></a>|<a href="https://pypi.org/project/ARCtrl/"><img alt="PyPI" src="https://img.shields.io/pepy/dt/arctrl?color=%234fb3d9"></a>|
 
 ## Install
 
-(currently only prereleases available, check the [nuget page](https://www.nuget.org/packages/ARCtrl) or [npm page](https://www.npmjs.com/package/@nfdi4plants/arctrl) respectively)
-
-### .NET
+#### .NET
 
 ```fsharp
 #r "nuget: ARCtrl"
@@ -23,12 +20,18 @@ and __JavaScript__! ❤️
 <PackageReference Include="ARCtrl" Version="1.1.0" />
 ```
 
-### JavaScript
+#### JavaScript
 
 ```bash
 npm i @nfdi4plants/arctrl
 ```
 
+#### Python
+
+```bash
+pip install arctrl
+```
+
 ## Docs
 
 Currently we provide some documentation in form of markdown files in the `/docs` folder of this repository!
@@ -37,7 +40,7 @@ Currently we provide some documentation in form of markdown files in the `/docs`
 
 ## Development
 
-### Requirements
+#### Requirements
 
 - [nodejs and npm](https://nodejs.org/en/download)
     - verify with `node --version` (Tested with v18.16.1)
@@ -47,7 +50,7 @@ Currently we provide some documentation in form of markdown files in the `/docs`
 - [Python](https://www.python.org/downloads/)
     - verify with `py --version` (Tested with 3.12.2, known to work only for >=3.11)
 
-### Local Setup
+#### Local Setup
 
 1. Setup dotnet tools
 
@@ -68,3 +71,18 @@ Currently we provide some documentation in form of markdown files in the `/docs`
    3. `.\.venv\Scripts\python.exe -m poetry install --no-root`
 
 Verify correct setup with `./build.cmd runtests` ✨
+
+## Performance
+
+Measured on 13th Gen Intel(R) Core(TM) i7-13800H
+
+| Name | Description | FSharp Time (ms) | JavaScript Time (ms) | Python Time (ms) |
+| --- | --- | --- | --- | --- |
+| Table_GetHashCode | From a table with 1 column and 10000 rows, retrieve the Hash Code |  5 | 21 | 226 |
+| Table_AddRows | Add 10000 rows to a table with 4 columns. |  15 | 22 | 289 |
+| Table_fillMissingCells | For a table 6 columns and 20000 rows, where each row has one missing value, fill those values with default values. | 49 | 108 | 4813 |
+| Table_ToJson | Serialize a table with 5 columns and 10000 rows to json. |  1099 | 481 | 6833 |
+| Table_ToCompressedJson | Serialize a table with 5 columns and 10000 rows to compressed json. |  261 |  2266 | 717334 |
+| Assay_toJson | Parse an assay with one table with 10000 rows and 6 columns to json |  915 | 2459 | 28799 |
+| Study_FromWorkbook | Parse a workbook with one study with 10000 rows and 6 columns to an ArcStudy |  97 | 87 | 1249 |
+| Investigation_ToWorkbook_ManyStudies | Parse an investigation with 1500 studies to a workbook |  621 | 379 | 9974 |

build/BasicTasks.fs

@@ -83,6 +83,17 @@ module Helper =
 
     let dotnet = createProcess "dotnet"
 
+    let node =
+        let nodePath =
+            match ProcessUtils.tryFindFileOnPath "node" with
+            | Some path -> path
+            | None ->
+                "node was not found in path. Please install it and make sure it's available from your path. " +
+                "See https://safe-stack.github.io/docs/quickstart/#install-pre-requisites for more info"
+                |> failwith
+
+        createProcess nodePath
+
     let npx =
         let npmPath =
             match ProcessUtils.tryFindFileOnPath "npx" with

build/Build.fs

@@ -30,6 +30,7 @@ let _preRelease =
         [setPrereleaseTag; clean; build; runTests; packPrerelease; createPrereleaseTag; publishNugetPrerelease; publishNPMPrerelease]
 
 ReleaseNotesTasks.updateReleaseNotes |> ignore
+PerformanceTasks.perforanceReport |> ignore
 
 [<EntryPoint>]
 let main args = 

build/Build.fsproj

@@ -13,6 +13,7 @@
     <Compile Include="BasicTasks.fs" />
     <Compile Include="GenerateIndexJs.fs" />
     <Compile Include="GenerateIndexPy.fs" />
+    <Compile Include="PerformanceTasks.fs" />
     <Compile Include="TestTasks.fs" />
     <Compile Include="PackageTasks.fs" />
     <Compile Include="ReleaseTasks.fs" />

build/PerformanceTasks.fs

@@ -0,0 +1,43 @@
+module PerformanceTasks
+
+open BlackFox.Fake
+open Fake.DotNet
+
+open ProjectInfo
+open BasicTasks
+open Fake.Core
+
+module PerformanceReport = 
+
+    let mutable cpu = ""
+
+    let promptCpu() =
+        if cpu = "" then
+            printfn "Please enter your cpu name"
+            cpu <- System.Console.ReadLine()
+
+    let testPerformancePy = BuildTask.create "testPerformancePy" [clean; build] {
+        promptCpu()
+        let path = "tests/Speedtest"
+     //transpile py files from fsharp code
+        run dotnet $"fable {path} -o {path}/py --lang python" ""
+        // run pyxpecto in target path to execute tests in python
+        run python $"{path}/py/program.py \"{cpu}\"" ""
+    }
+    let testPerformanceJs = BuildTask.create "testPerformanceJS" [clean; build] {
+        promptCpu()
+        let path = "tests/Speedtest"
+        // transpile js files from fsharp code
+        run dotnet $"fable {path} -o {path}/js" ""
+        // run mocha in target path to execute tests
+        run node $"{path}/js/program.js \"{cpu}\"" ""
+    }
+    let testPerformanceDotnet = BuildTask.create "testPerformanceDotnet" [clean; build] {
+        promptCpu()
+        let path = "tests/Speedtest"
+        run dotnet $"run --project {path} \"{cpu}\"" ""
+    }
+
+let perforanceReport = BuildTask.create "PerformanceReport" [PerformanceReport.testPerformancePy; PerformanceReport.testPerformanceJs; PerformanceReport.testPerformanceDotnet] { 
+    ()
+}

docs/scripts_fsharp/ArcInvestigation.fsx

@@ -20,7 +20,7 @@ open FsSpreadsheet.Net
 
 let fswb = ArcInvestigation.toFsWorkbook investigation_comments
 
-fswb.ToFile("test2.isa.investigation.xlsx")
+fswb.ToXlsxFile("test2.isa.investigation.xlsx")
 
 // ## Json - Write
 

tests/ISA/ISA.Json.Tests/ArcTypes.Tests.fs

@@ -179,16 +179,6 @@ let tests_ArcTable = testList "ArcTable" [
       let decoded = ArcTable.fromCompressedJsonString encoded
       Expect.equal decoded filled "empty table is wrong after compressed encoding and decoding"
     // Set to pTest in Fable, as compressed parsing is around 10times slower than uncompressed parsing. This is weird, since in dotnet it is around 10times faster
-    #if FABLE_COMPILER 
-    ptestCase "Performance" <| fun _ ->
-    #else
-    testCase "Performance" <| fun _ ->
-    #endif
-        let t = TestObjects.Spreadsheet.Study.LargeFile.table
-        Expect.isFasterThan (t.ToCompressedJsonString >> ignore) (t.ToJsonString  >> ignore) "toCompressedJsonString is slower than to uncompressed"
-        let json = t.ToJsonString()
-        let compressed = Expect.wantFaster (t.ToCompressedJsonString) 1000 "toCompressedJsonString should be faster"
-        Expect.isTrue (compressed.Length*5 < json.Length) $"compressed should be more than 10 times smaller than uncompressed, but was only {float json.Length / float compressed.Length}x smaller"
     testCase "rangeColumnSize" <| fun _ ->
         // testTable column should be saved as range column, this should make it smaller than the IO column even though it has more cells
         let testTable = ArcTable.init("Test")
@@ -258,44 +248,6 @@ let tests_ArcAssay = testList "ArcAssay" [
       let decoded = ArcAssay.fromCompressedJsonString encoded
       Expect.equal decoded filled "empty table is wrong after compressed encoding and decoding"
   ]
-  testList "performance" [
-    testCase "SingleLongTable_JsonAssay" <| fun _ ->
-        let a = ArcAssay.init("MyAssay")
-        let t = a.InitTable("MyTable")
-        t.AddColumn(CompositeHeader.Input IOType.Source)
-        t.AddColumn(CompositeHeader.Parameter (OntologyAnnotation.fromString("MyParameter1")))
-        t.AddColumn(CompositeHeader.Parameter (OntologyAnnotation.fromString("MyParameter2")))
-        t.AddColumn(CompositeHeader.Parameter (OntologyAnnotation.fromString("MyParameter3")))
-        t.AddColumn(CompositeHeader.Characteristic (OntologyAnnotation.fromString("MyCharacteristic")))
-        t.AddColumn(CompositeHeader.Output IOType.Sample)
-        let rowCount = 10000
-        for i = 0 to rowCount - 1 do
-            let cells =             
-                [|
-                    CompositeCell.FreeText $"Source{i}"
-                    CompositeCell.FreeText $"Parameter1_value"
-                    CompositeCell.FreeText $"Parameter2_value"
-                    CompositeCell.FreeText $"Parameter3_value{i - i % 10}"
-                    CompositeCell.FreeText $"Characteristic_value"
-                    CompositeCell.FreeText $"Sample{i}"
-                |]
-            for j = 0 to cells.Length - 1 do
-                t.Values.[(j,i)] <- cells.[j]
-        let f() = ArcAssay.toJsonString a
-        #if FABLE_COMPILER_JAVASCRIPT
-        let expectedMs = 5000
-        #endif
-        #if FABLE_COMPILER_PYTHON
-        let expectedMs = 100000
-        #endif
-        #if !FABLE_COMPILER
-        let expectedMs = 2500
-        #endif
-        // 1200ms in Dotnet on i7-13800H
-        // 3412ms in Javascript on i7-13800H
-        // 24562ms in Javascript on i7-13800H
-        Expect.wantFaster f expectedMs "toJsonString should be faster" |> ignore
-  ]
 ]
 
 let tests_ArcStudy = testList "ArcStudy" [

tests/ISA/ISA.Spreadsheet.Tests/ISA.Spreadsheet.Tests.fsproj

@@ -17,7 +17,6 @@
     <Compile Include="StudyFileTests.fs" />
     <Compile Include="AssayFileTests.fs" />
     <Compile Include="InvestigationFileTests.fs" />
-    <Compile Include="Performance.Tests.fs" />
     <Compile Include="Main.fs" />
   </ItemGroup>
   <ItemGroup>

tests/ISA/ISA.Spreadsheet.Tests/Main.fs

@@ -12,7 +12,6 @@ let all = testSequenced <| testList "ISA.Spreadsheet" [
     ArcStudyTests.main
     SparseTableTests.main
     IdentifierTests.main
-    Tests.Performance.Main
 ]
 
 [<EntryPoint>]