Leiden algorithm included in precompiled macOS binaries

refresh-bio · Oct 13, 2024 · 4e2d935 · 4e2d935
1 parent 5503c50
commit 4e2d935
Show file tree

Hide file tree

Showing 6 changed files with 87 additions and 151 deletions.
diff --git a/.github/workflows/deploy.yml b/.github/workflows/deploy.yml
@@ -17,6 +17,8 @@ jobs:
     runs-on: [self-hosted, clusty, '${{ matrix.machine }}']
 
     steps:
+    - name: clean
+      run: rm -rf ${{ github.workspace }}/*
     - uses: actions/checkout@v4
       with:
         submodules: recursive
@@ -36,30 +38,18 @@ jobs:
       fail-fast: false
       matrix:
         machine: [x64_linux]
-        leiden: [true]
         platform: [avx2]
-        compiler: [g++-13]
+        compiler: [14]
         include:
-        - machine: arm64_linux
-          leiden: true
-          platform: arm8
-          compiler: g++-12
-        - machine: x64_mac
-          leiden: false
-          platform: avx2
-          compiler: g++-13
-        - machine: arm64_mac
-          leiden: false
-          platform: m1
-          compiler: g++-13
+        - {machine: arm64_linux, platform: arm8, compiler: 12}
+        - {machine: x64_mac, platform: avx2, compiler: 12}
+        - {machine: arm64_mac, platform: m1, compiler: 12}
 
     runs-on: [self-hosted, clusty, '${{ matrix.machine }}']  
 
     steps:
     - name: make 
-      run: |
-        make clean
-        make -j32 CXX=${{matrix.compiler}} STATIC_LINK=true LEIDEN=${{ matrix.leiden }} PLATFORM=${{ matrix.platform }}
+      run: make -j32 CXX=g++-${{matrix.compiler}} CC=gcc-${{matrix.compiler}} PLATFORM=${{ matrix.platform }} LEIDEN=true
     - name: tar artifacts
       run: tar -cvzf clusty.tar.gz clusty LICENSE
 

diff --git a/.github/workflows/self-hosted.yml b/.github/workflows/self-hosted.yml
@@ -19,6 +19,8 @@ jobs:
     runs-on: [self-hosted, clusty, '${{ matrix.machine }}']
 
     steps:
+    - name: clean
+      run: rm -rf ${{ github.workspace }}/*
     - uses: actions/checkout@v4
       with:
         submodules: recursive
@@ -39,26 +41,20 @@ jobs:
       fail-fast: false
       matrix:
         machine: [x64_linux, x64_mac, arm64_linux, arm64_mac]
-        compiler: [g++-11, g++-12, g++-13]
+        compiler: [11, 12]
         include:
-        - machine: x64_linux
-          platform: avx2
-        - machine: arm64_linux
-          platform: arm8
-        - machine: x64_mac
-          platform: avx2
-        - machine: arm64_mac
-          platform: m1
-        exclude:
-        - machine: arm64_linux
-          compiler: g++-13
-
+        - {machine: x64_linux, platform: avx2, compiler: 14}
+        - {machine: x64_linux, platform: avx2}
+        - {machine: arm64_linux, platform: arm8}
+        - {machine: x64_mac, platform: avx2}
+        - {machine: arm64_mac, platform: m1}
+
     runs-on: [self-hosted, clusty, '${{ matrix.machine }}']
 
     steps:
     - name: make
       run: | 
-        make -j32 CXX=${{matrix.compiler}} PLATFORM=${{ matrix.platform }}
+        make -j32 CXX=g++-${{matrix.compiler}} CC=gcc-${{matrix.compiler}} PLATFORM=${{ matrix.platform }} LEIDEN=true
         cp ./clusty ./clusty-${{matrix.compiler}}
         make clean
 
@@ -69,13 +65,12 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        machine: [x64_linux, x64_mac, arm64_linux, arm64_mac]
-        compiler: [g++-11, g++-12, g++-13]
         algo: [single]
-        exclude:
-        - machine: arm64_linux
-          compiler: g++-13
-
+        machine: [x64_linux, x64_mac, arm64_linux, arm64_mac]
+        compiler: [11, 12]
+        include:
+        - {algo: single, machine: x64_linux, compiler: 14}
+
     runs-on: [self-hosted, clusty, '${{ matrix.machine }}']    
 
     steps:
@@ -88,18 +83,20 @@ jobs:
 
   vir61-linkage:
     name: Vir61
-    needs: make
+    needs: toy
     strategy:
       fail-fast: false
       matrix:
-        machine: [x64_linux, x64_mac, arm64_linux, arm64_mac]
-        compiler: [g++-11, g++-12, g++-13]
         algo: [single, complete]
         threshold: [95, 70]
-        exclude:
-        - machine: arm64_linux
-          compiler: g++-13
-
+        machine: [x64_linux, x64_mac, arm64_linux, arm64_mac]
+        compiler: [12]
+        include:
+        - {algo: single, threshold: 95, machine: x64_linux, compiler: 14}
+        - {algo: single, threshold: 70, machine: x64_linux, compiler: 14}
+        - {algo: complete, threshold: 95, machine: x64_linux, compiler: 14}
+        - {algo: complete, threshold: 70, machine: x64_linux, compiler: 14}
+
     runs-on: [self-hosted, clusty, '${{ matrix.machine }}']    
 
     steps:
@@ -111,18 +108,20 @@ jobs:
   ########################################################################################   
   ictv-linkage:
     name: ICTV (single and complete versus python clusterings)
-    needs: make
+    needs: vir61-linkage
     strategy:
       fail-fast: false
       matrix:
-        machine: [x64_linux, x64_mac, arm64_linux, arm64_mac]
-        compiler: [g++-11, g++-12, g++-13]
         algo: [single, complete]
         threshold: [95, 70]
-        exclude:
-        - machine: arm64_linux
-          compiler: g++-13
-
+        machine: [x64_linux, x64_mac, arm64_linux, arm64_mac]
+        compiler: [12]
+        include:
+        - {algo: single, threshold: 95, machine: x64_linux, compiler: 14}
+        - {algo: single, threshold: 70, machine: x64_linux, compiler: 14}
+        - {algo: complete, threshold: 95, machine: x64_linux, compiler: 14}
+        - {algo: complete, threshold: 70, machine: x64_linux, compiler: 14}
+
     runs-on: [self-hosted, clusty, '${{ matrix.machine }}']    
 
     steps:
@@ -149,18 +148,22 @@ jobs:
   ########################################################################################   
   ictv-all:
     name: ICTV (all algos versus our clusterings)
-    needs: make
+    needs: ictv-linkage
     strategy:
       fail-fast: false
       matrix:
-        machine: [x64_linux, x64_mac, arm64_linux, arm64_mac]
-        compiler: [g++-11, g++-12, g++-13]
-        algo: [single, complete, uclust, set-cover, cd-hit]
+        algo: [single, complete, uclust, set-cover, cd-hit, leiden]
         threshold: [70]
-        exclude:
-        - machine: arm64_linux
-          compiler: g++-13
-
+        machine: [x64_linux, x64_mac, arm64_linux, arm64_mac]
+        compiler: [12]
+        include:
+        - {algo: single, threshold: 70, machine: x64_linux, compiler: 14}
+        - {algo: complete, threshold: 70, machine: x64_linux, compiler: 14}
+        - {algo: uclust, threshold: 70, machine: x64_linux, compiler: 14}
+        - {algo: set-cover, threshold: 70, machine: x64_linux, compiler: 14}
+        - {algo: cd-hit, threshold: 70, machine: x64_linux, compiler: 14}
+        - {algo: leiden, threshold: 70, machine: x64_linux, compiler: 14}
+
     runs-on: [self-hosted, clusty, '${{ matrix.machine }}']    
 
     steps:
@@ -178,69 +181,3 @@ jobs:
       run: |
         ./clusty-${{matrix.compiler}} --algo ${{matrix.algo}} --objects-file ./test/ictv.list --similarity --min ani 0.${{matrix.threshold}} ./test/ictv.num ictv.${{matrix.algo}}.${{matrix.threshold}}.reps.num --out-representatives --numeric-ids
         cmp ictv.${{matrix.algo}}.${{matrix.threshold}}.reps.num ./test/ictv.${{matrix.algo}}.${{matrix.threshold}}.reps.csv
-
-
- ########################################################################################
-  make-leiden:
-    name: Make (Leiden)
-
-    needs: ictv-all
-    strategy:
-      fail-fast: false
-      matrix:
-        machine: [x64_linux, arm64_linux]
-        include:
-        - machine: x64_linux
-          compiler: g++-13
-          platform: avx2
-        - machine: arm64_linux
-          compiler: g++-12
-          platform: arm8
-
-    runs-on: [self-hosted, clusty, '${{ matrix.machine }}']
-
-    steps:
-    - name: make
-      run: | 
-        make -j32 CXX=${{matrix.compiler}} LEIDEN=true PLATFORM=${{ matrix.platform }}
-        cp ./clusty ./clusty-leiden-${{matrix.compiler}}
-        make clean
-        
-  ########################################################################################   
-  ictv-leiden:
-    name: ICTV (Leiden)
-    needs: make-leiden
-    strategy:
-      fail-fast: false
-      matrix:
-        machine: [x64_linux, arm64_linux]
-        algo: [leiden]
-        threshold: [70]
-        include:
-        - machine: x64_linux
-          compiler: g++-13
-        - machine: arm64_linux
-          compiler: g++-12
-
-    runs-on: [self-hosted, clusty, '${{ matrix.machine }}']    
-
-    steps:
-    - name: ${{matrix.algo}}, ${{matrix.threshold}} (no representatives)
-      run: |
-        ./clusty-leiden-${{matrix.compiler}} --algo ${{matrix.algo}} --objects-file ./test/ictv.list --similarity --min ani 0.${{matrix.threshold}} ./test/ictv.ani ictv.${{matrix.algo}}.${{matrix.threshold}}
-        cmp ictv.${{matrix.algo}}.${{matrix.threshold}} ./test/ictv.${{matrix.algo}}.${{matrix.threshold}}.csv
-        
-    - name: ${{matrix.algo}}, ${{matrix.threshold}} (with representatives)
-      run: |
-        ./clusty-leiden-${{matrix.compiler}} --algo ${{matrix.algo}} --objects-file ./test/ictv.list --similarity --min ani 0.${{matrix.threshold}} ./test/ictv.ani ictv.${{matrix.algo}}.${{matrix.threshold}}.reps --out-representatives
-        cmp ictv.${{matrix.algo}}.${{matrix.threshold}}.reps ./test/ictv.${{matrix.algo}}.${{matrix.threshold}}.reps.csv
-        
-    - name: ${{matrix.algo}}, ${{matrix.threshold}} (with representatives, numeric ids)
-      run: |
-        ./clusty-leiden-${{matrix.compiler}} --algo ${{matrix.algo}} --objects-file ./test/ictv.list --similarity --min ani 0.${{matrix.threshold}} ./test/ictv.num ictv.${{matrix.algo}}.${{matrix.threshold}}.reps.num --out-representatives --numeric-ids
-        cmp ictv.${{matrix.algo}}.${{matrix.threshold}}.reps.num ./test/ictv.${{matrix.algo}}.${{matrix.threshold}}.reps.csv
-     
-    - name: ${{matrix.algo}}, ${{matrix.threshold}} (non-default params)
-      run: |
-        ./clusty-leiden-${{matrix.compiler}} --algo ${{matrix.algo}} --objects-file ./test/ictv.list --similarity --min ani 0.${{matrix.threshold}} --leiden-resolution 1.0 --leiden-beta 0.03 --leiden-iterations 3 ./test/ictv.ani ictv.${{matrix.algo}}-params.${{matrix.threshold}}.reps.csv --out-representatives 
-        cmp ictv.${{matrix.algo}}-params.${{matrix.threshold}}.reps.csv ./test/ictv.${{matrix.algo}}-params.${{matrix.threshold}}.reps.csv
diff --git a/README.md b/README.md
@@ -2,7 +2,7 @@
 
 [![Bioconda downloads](https://img.shields.io/conda/dn/bioconda/clusty.svg?style=flag&label=Bioconda%20downloads)](https://anaconda.org/bioconda/clusty)
 [![GitHub downloads](https://img.shields.io/github/downloads/refresh-bio/clusty/total.svg?style=flag&label=GitHub%20downloads)](https://github.com/refresh-bio/clusty/releases)
-[![GitHub Actions CI](../../workflows/GitHub%20Actions%20CI/badge.svg)](../../actions/workflows/main.yml)
+[![Build and tests](../../workflows/GitHub%20Actions%20CI/badge.svg)](../../actions/workflows/main.yml)
 [![License](https://anaconda.org/bioconda/famsa/badges/license.svg)](https://www.gnu.org/licenses/gpl-3.0.html)
 
 ![x86-64](https://img.shields.io/static/v1?label=%E2%80%8B&message=x86-64&color=yellow&logo=PCGamingWiki&logoColor=white)
@@ -43,7 +43,7 @@ cd ./test
 
 ## Installation
 
-Clusty comes with a set of precompiled binaries for Windows and Linux. They can be found under [Releases](./releases) tab.
+Clusty comes with a set of precompiled binaries for Windows, Linux, and macOS. They can be found under [Releases](./releases) tab.
 The software is also available on [Bioconda](https://anaconda.org/bioconda/clusty):
 ```
 conda install -c bioconda clusty
@@ -52,9 +52,16 @@ For detailed instructions how to set up Bioconda, please refer to the [Bioconda
 
 The package can be built from the sources distributed as:
 * Visual Studio 2022 solution for Windows,
-* MAKE project for Linux (g++-10 required).
+* MAKE project for Linux and macOS (g++-10 required).
 
-Clusty provides igraph's implementation of the Leiden algorithm. However, as igraph requires several external dependencies (CMake 3.18, Flex, Bison), it is by default not linked to the Clusty software. To install dependencies under Debian/Ubuntu linux use the following command:
+To compile Clusty under Linux/macOS please run:
+```
+make -j 
+```
+
+### Leiden algorithm
+
+Clusty provides igraph's implementation of the Leiden algorithm. Precompiled binaries as well as bioconda distributions include Leiden algorithm. However, as igraph requires several external dependencies (CMake 3.18, Flex, Bison), it is by default not linked to the Clusty software. To install dependencies under Debian/Ubuntu linux use the following command:
 ```
 sudo apt-get install cmake flex bison
 ```

diff --git a/makefile b/makefile
@@ -67,17 +67,17 @@ else
 endif
 
 
-ifeq ($(STATIC_LINK), true) 
+ifeq ($(DYNAMIC_LINK), true)
+    CFLAGS	= -Wall -O3 $(ARCH_FLAGS) -std=c++17 $(DEFINE_FLAGS) $(INCLUDES) -pthread 
+    CLINK	= -lm -O3 -std=c++17 -pthread $(ABI_FLAGS) 
+else
 ifeq ($(uname_S),Darwin)
     CFLAGS  = -Wall -O3 $(ARCH_FLAGS) -std=c++17 $(DEFINE_FLAGS) $(INCLUDES)
     CLINK	= -lm -O3 -std=c++17 $(ABI_FLAGS) -static-libgcc
 else
     CFLAGS  = -Wall -O3 $(ARCH_FLAGS) -std=c++17 $(DEFINE_FLAGS) $(INCLUDES) -static -Wl,--whole-archive -lpthread -Wl,--no-whole-archive
     CLINK	= -lm -static -O3 -std=c++17 $(ABI_FLAGS) -Wl,--whole-archive -lpthread -Wl,--no-whole-archive
-endif
-else
-    CFLAGS	= -Wall -O3 $(ARCH_FLAGS) -std=c++17 $(DEFINE_FLAGS) $(INCLUDES) -pthread 
-    CLINK	= -lm -O3 -std=c++17 -pthread $(ABI_FLAGS) 
+endif  
 endif
 
 ifeq ($(LEIDEN), true)  

diff --git a/src/params.cpp b/src/params.cpp
@@ -47,7 +47,7 @@ void Params::printUsage() const {
 		<< endl << endl
 		<< "  " + PARAM_LEIDEN_RESOLUTION + " - resolution parameter for Leiden algorithm (default: " << leidenParams.resolution << ")" << endl
 		<< "  " + PARAM_LEIDEN_BETA + " - beta parameter for Leiden algorithm (default: " << leidenParams.beta << ")" << endl
-		<< "  " + PARAM_LEIDEN_RESOLUTION + " - number of interations for Leiden algorithm (default: " << leidenParams.numIterations << ")"
+		<< "  " + PARAM_LEIDEN_ITERATIONS + " - number of interations for Leiden algorithm (default: " << leidenParams.numIterations << ")"
 #endif
 		<< endl << endl;
 }

diff --git a/src/version.h b/src/version.h
@@ -8,33 +8,35 @@
 //
 // *******************************************************************************************
 
-#define VERSION		"1.1.1"
-#define DATE		"2024-10-05"
+#define VERSION		"1.1.2"
+#define DATE		"2024-10-13"
 
 
 /********* Version history *********
 
- 1.1.1 (2024-10-05)
+1.1.2
+* Precompiled binaries for macOS include Leiden algorithm.
+* Fixed small bug with `--leiden-iterations` param being displayed in help as `--leiden-resolution`.
+
+1.1.1 (2024-10-05)
 Fixed bug with Leiden algorithm always running with default parameters.
 
- 1.1.0 (2024-09-30)
- Memory optimizatons:
+1.1.0 (2024-09-30)
+Memory optimizatons:
 * Row identifier removed from dist_t structure.
 * Object identifiers in complete linkage stored as 32-bit integers.
 * For some clustering algorithms only edges are stored (without distances). 
 
- 1.0.3 (2024-09-12)
- Fixes in building scripts.
-
- 1.0.2 (2024-08-29)
- Fixed crash in reading very large input matrices. Reduced memory requirements.
-
+1.0.3 (2024-09-12)
+Fixes in building scripts.
 
- 1.0.1 (2024-08-19)
- Fixed bug in parsing floating point numbers with more than 15 decimal places.
+1.0.2 (2024-08-29)
+Fixed crash in reading very large input matrices. Reduced memory requirements.
 
+1.0.1 (2024-08-19)
+Fixed bug in parsing floating point numbers with more than 15 decimal places.
 
- 1.0.0 (2024-06-26)
- First public release
+1.0.0 (2024-06-26)
+First public release
 
- */
+*/