add docs

OnRunning.md

@@ -0,0 +1,27 @@
+# Running Experiments with Lambda:
+This is not supposed to be a instruction on how to do it properly, but it is a writeup on how i did it.
+If you want to do it properly, extend the command line Arguments for haga-lambda and allow runtime tweaking of Hyperparams and Datasets. While at it, generalizing LamdaCalculusV1 would be smart, too. You can use LamdaCalculusV2 as a template on how to do it more properly. (I wrote that later, and was IMO quite a bit smarter about it. I sadly didn't have time to fix up V1...)
+
+You just want to do the same hack i did or know about it?
+
+create a branch for each Dataset-experiment pair. e.g. iris_1 ... iris_9
+
+here git is your friend, especially if you inevitably screw up.
+e.g. echo git\ checkout\ iris_{1..9}\;\ git\ cherry-pick\ 7ced1e1\; will create a command for applying the commit 7ced1e1 to every iris branch.
+
+Adapt the build.sbatch and run.sbatch and **commit them**!
+clone the branch you committed to on the cluster.
+create the required folders! If you forget the output one, slurm will fail silently!
+
+Make sure to sbatch an adapted **build.sbatch before run.sbatch**!
+build.sbatch will need to be adapted for and run on every node you will use!
+Otherwise stuff WILL break!
+
+sbatch run.sbatch
+
+You can use squeue to monitor progress.
+
+A huge slew of raw data will be dumped into the output Folder. The error files contain results, the output files stats during training.
+
+On how to process these results, see: https://merl.dnshome.de/git/Hans/haga-graphics
+

build.sbatch

@@ -1,9 +1,28 @@
 #!/usr/bin/env bash
 #SBATCH --time=00:10:00
 #SBATCH --partition=cpu
-#SBATCH --output=./output/output_build.txt
-#SBATCH --error=./output/error_build.txt
+# 9 Experiments * 3 Datasets
+#SBATCH --array=0-27
+# ensure output exists, is a folder and is writable in your working directory
+#SBATCH --output=./output/output_run_%a.txt
+#SBATCH --error=./output/error_run_%a.txt
+# run once for every node you plan to use
 #SBATCH --nodelist=oc-compute02
-#SBATCH --mem=4G
-#SBATCH -c16
-srun nix develop --command stack --no-nix --system-ghc --no-install-ghc build
+#SBATCH --mem=2G
+
+
+# list your branches
+problems=("iris" "nurse" "german")
+
+#9 Experiments
+current_problem=${problems[(${SLURM_ARRAY_TASK_ID}/9)]}
+#9 Experiments
+current_variant=$(((${SLURM_ARRAY_TASK_ID}) % 9 + 1))
+current_branch="${current_problem}_${current_variant}"
+
+# ensure [full path to writable folder on node *] exists
+git clone -b $current_branch --single-branch "[your git repo]" [full path to writable folder on node 1]/$current_branch
+git clone -b $current_branch --single-branch "[your git repo]" [full path to writable folder on node 1]/$current_branch
+#... for every node
+
+srun bash -c "cd /data/$SLURMD_NODENAME/merljoha/$current_branch; nix develop --command stack --no-nix --system-ghc --no-install-ghc build"

lambda/src/LambdaDatasets/GermanDataset.hs

@@ -38,7 +38,7 @@ lE =
             ((Ref.SomeTypeRep (Ref.TypeRep @(Int -> Int -> Int))), ["(+)", "(-)", "(*)"]),
             -- Logic
             ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> Bool -> Bool))), ["(&&)", "(||)"]),
-            -- Ordered Enums
+            -- Ordered
             ((Ref.SomeTypeRep (Ref.TypeRep @(Int -> Int -> Bool))), ["(>)", "(==)", "(/=)", "(>=)"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(AccountStatus -> AccountStatus -> Bool))), ["(>)", "(==)", "(/=)", "(>=)"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(CreditHistory -> CreditHistory -> Bool))), ["(>)", "(==)", "(/=)", "(>=)"]),
@@ -46,7 +46,7 @@ lE =
             ((Ref.SomeTypeRep (Ref.TypeRep @(EmploymentStatus -> EmploymentStatus -> Bool))), ["(>)", "(==)", "(/=)", "(>=)"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(OtherDebtors -> OtherDebtors -> Bool))), ["(>)", "(==)", "(/=)", "(>=)"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(Job -> Job -> Bool))), ["(>)", "(==)", "(/=)", "(>=)"]),
-            -- Eq Enum
+            -- Eq
             ((Ref.SomeTypeRep (Ref.TypeRep @(GermanClass -> GermanClass -> Bool))), ["(==)", "(/=)"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(Purpose -> Purpose -> Bool))), ["(==)", "(/=)"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(StatusAndSex -> StatusAndSex -> Bool))), ["(==)", "(/=)"]),
@@ -55,7 +55,7 @@ lE =
             ((Ref.SomeTypeRep (Ref.TypeRep @(Housing -> Housing -> Bool))), ["(==)", "(/=)"]),
             -- Any Type
             ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> Int -> Int -> Int))), ["if'"]),
-            ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> GermanClass -> GermanClass -> GermanClass))), ["if'","if'","if'","if'","if'","if'","if'","if'"]),
+            ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> GermanClass -> GermanClass -> GermanClass))), ["if'"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> AccountStatus -> AccountStatus -> AccountStatus))), ["if'"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> CreditHistory -> CreditHistory -> CreditHistory))), ["if'"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> Purpose -> Purpose -> Purpose))), ["if'"]),
@@ -86,13 +86,13 @@ lE =
             ((Ref.SomeTypeRep (Ref.TypeRep @(Job))), [(fmap show (enumUniform UnemployedOrUnskilledNonResident HighlySkilled ))])
           ],
       targetType = (Ref.SomeTypeRep (Ref.TypeRep @(AccountStatus -> Int -> CreditHistory -> Purpose -> Int -> Savings -> EmploymentStatus -> Int -> StatusAndSex -> OtherDebtors -> Int -> Property -> Int -> OtherPlans -> Housing -> Int -> Job -> Int -> Bool -> Bool -> GermanClass))),
-      maxDepth = 8,
+      maxDepth = 9,
       weights =
         ExpressionWeights
-          { lambdaSpucker = 1,
-            lambdaSchlucker = 2,
-            symbol = 30,
-            variable = 10,
+          { lambdaSpucker = 0,
+            lambdaSchlucker = 10,
+            symbol = 100,
+            variable = 5,
             constant = 5
           }
     }

lambda/src/LambdaDatasets/IrisDataset.hs

@@ -34,12 +34,17 @@ lE =
   LambdaEnviroment
     { functions =
         Map.fromList
-          [ ((Ref.SomeTypeRep (Ref.TypeRep @(Float -> Float -> Float))), ["(+)", "(-)", "(*)"]),
-            ((Ref.SomeTypeRep (Ref.TypeRep @(Float -> Float -> Bool))), ["(>)", "(==)", "(>=)"]),
-            ((Ref.SomeTypeRep (Ref.TypeRep @(IrisClass -> IrisClass -> Bool))), ["(==)"]),
-            ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> Float -> Float -> Float))), ["if'","if'","if'","if'","if'","if'","if'","if'"]),
+          [  -- Math
+            ((Ref.SomeTypeRep (Ref.TypeRep @(Float -> Float -> Float))), ["(+)", "(-)", "(*)"]),
+            -- Logic
             ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> Bool -> Bool))), ["(&&)", "(||)"]),
-            ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> IrisClass -> IrisClass -> IrisClass))), ["if'","if'","if'","if'","if'","if'","if'","if'","if'","if'"])
+            -- Ordered
+            ((Ref.SomeTypeRep (Ref.TypeRep @(Float -> Float -> Bool))), ["(>)", "(==)", "(/=)", "(>=)"]),
+            -- Eq
+            ((Ref.SomeTypeRep (Ref.TypeRep @(IrisClass -> IrisClass -> Bool))), ["(==)","(/=)"]),
+            -- Any Type
+            ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> Float -> Float -> Float))), ["if'"]),
+            ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> IrisClass -> IrisClass -> IrisClass))), ["if'"])
           ],
       constants =
         Map.fromList
@@ -48,13 +53,13 @@ lE =
             ((Ref.SomeTypeRep (Ref.TypeRep @(IrisClass))), [(fmap show (enumUniform Setosa Versicolor :: RVar IrisClass))])
           ],
       targetType = (Ref.SomeTypeRep (Ref.TypeRep @(Float -> Float -> Float -> Float -> IrisClass))),
-      maxDepth = 10,
+      maxDepth = 9,
       weights =
         ExpressionWeights
-          { lambdaSpucker = 1,
-            lambdaSchlucker = 1,
-            symbol = 30,
-            variable = 100,
+          { lambdaSpucker = 0,
+            lambdaSchlucker = 10,
+            symbol = 100,
+            variable = 5,
             constant = 5
           }
     }
@@ -63,7 +68,7 @@ lEE :: LamdaExecutionEnv
 lEE =
   LamdaExecutionEnv
     { -- For now these need to define all available functions and types. Generic functions can be used.
-      imports = ["LambdaDatasets.IrisDataset"],
+      imports = ["LambdaDatasets.IrisDefinition"],
       training = True,
       trainingData =
         ( map fst (takeFraktion 0.8 irisTrainingData),
@@ -84,7 +89,7 @@ shuffledLEE = do
   itD <- smpl $ shuffle irisTrainingData
   return  LamdaExecutionEnv
     { -- For now these need to define all available functions and types. Generic functions can be used.
-      imports = ["LambdaDatasets.IrisDataset"],
+      imports = ["LambdaDatasets.IrisDefinition"],
       training = True,
       trainingData =
         ( map fst (takeFraktion 0.8 itD),

lambda/src/LambdaDatasets/NurseryDataset.hs

@@ -37,7 +37,7 @@ lE =
           [ -- Math
             -- Logic
             ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> Bool -> Bool))), ["(&&)", "(||)"]),
-            -- Ordered Enums
+            -- Ordered
             ((Ref.SomeTypeRep (Ref.TypeRep @(NurseryClass -> NurseryClass -> Bool))), ["(>)", "(==)", "(/=)", "(>=)"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(Parents -> Parents -> Bool))), ["(>)", "(==)", "(/=)", "(>=)"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(HasNurs -> HasNurs -> Bool))), ["(>)", "(==)", "(/=)", "(>=)"]),
@@ -47,7 +47,7 @@ lE =
             ((Ref.SomeTypeRep (Ref.TypeRep @(Finance -> Finance -> Bool))), ["(>)", "(==)", "(/=)", "(>=)"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(Social -> Social -> Bool))), ["(>)", "(==)", "(/=)", "(>=)"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(Health -> Health -> Bool))), ["(>)", "(==)", "(/=)", "(>=)"]),
-            -- Eq Enum
+            -- Eq
             -- Any Type
             ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> Int -> Int -> Int))), ["if'"]),
             ((Ref.SomeTypeRep (Ref.TypeRep @(Bool -> NurseryClass -> NurseryClass -> NurseryClass))), ["if'","if'","if'","if'","if'","if'","if'","if'"]),
@@ -74,13 +74,13 @@ lE =
             ((Ref.SomeTypeRep (Ref.TypeRep @(Health))), [(fmap show (enumUniform NotRecommendHealth PriorityHealth ))])
           ],
       targetType = (Ref.SomeTypeRep (Ref.TypeRep @(Parents -> HasNurs -> Form -> Children -> Housing -> Finance -> Social -> Health -> NurseryClass))),
-      maxDepth = 8,
+      maxDepth = 9,
       weights =
         ExpressionWeights
-          { lambdaSpucker = 1,
-            lambdaSchlucker = 2,
-            symbol = 30,
-            variable = 20,
+          { lambdaSpucker = 0,
+            lambdaSchlucker = 10,
+            symbol = 100,
+            variable = 5,
             constant = 5
           }
     }

lambda/src/Main.hs

@@ -8,8 +8,8 @@ import Pipes
 import Pretty
 import Protolude hiding (for)
 import System.IO
--- import LambdaDatasets.IrisDataset
-import LambdaDatasets.NurseryDataset
+import LambdaDatasets.IrisDataset
+-- import LambdaDatasets.NurseryDataset
 -- import LambdaDatasets.GermanDataset
 import Debug.Trace as DB
 import qualified Data.Map.Strict as Map
@@ -35,7 +35,7 @@ options =
       ( long "population-size"
           <> short 'p'
           <> metavar "N"
-          <> value 400
+          <> value 100
           <> help "Population size"
       )
 
@@ -59,7 +59,7 @@ main =
       selectionType = Tournament 3,
       termination = (steps (iterations opts)),
       poulationSize = (populationSize opts),
-      stepSize = 120,
+      stepSize = 90,
       elitismRatio = 5/100
     }
     pop' <- runEffect (for (run cfg) logCsv)

run.sbatch

@@ -1,9 +1,31 @@
 #!/usr/bin/env bash
-#SBATCH --time=18:00:00
+# test this timing, it scales with result sizes
+#SBATCH --time=12:00:00
 #SBATCH --partition=cpu
-#SBATCH --array=0-30
-#SBATCH --output=./output/output_run_%j.txt
-#SBATCH --error=./output/error_run_%j.txt
-#SBATCH --nodelist=oc-compute02
-#SBATCH --mem=3G
-srun nix develop --command stack --no-nix --system-ghc --no-install-ghc run haga-lambda
+# 30 Runs * 9 Experiments * 3 Datasets
+#SBATCH --array=0-809
+# ensure output exists, is a folder and is writable in your working directory
+#SBATCH --output=./output/output_run_%a.txt
+#SBATCH --error=./output/error_run_%a.txt
+# exclude nodes with weaker CPUs
+#SBATCH --exclude=oc222
+# test memory usage, it scales **Exponentially** with max Depth. Implement some countermeasures if that's a problem, e.g. raise max depth over time.
+#SBATCH --mem=6G
+#SBATCH --nodes=1
+
+
+# list your branches
+problems=("iris" "nurse" "german")
+
+# 30 Runs * 9 Experiments
+current_problem=${problems[(${SLURM_ARRAY_TASK_ID}/270)]}
+# 30 Runs, 9 Experiments
+current_variant=$(((${SLURM_ARRAY_TASK_ID} / 30) % 9 + 1))
+current_branch="${current_problem}_${current_variant}"
+
+# ensure [full path to writable folder on node *] exists
+git clone -b $current_branch --single-branch "[your git repo]" [full path to writable folder on node 1]/$current_branch
+git clone -b $current_branch --single-branch "[your git repo]" [full path to writable folder on node 2]/$current_branch
+#... for every node
+
+srun bash -c "cd /data/$SLURMD_NODENAME/merljoha/$current_branch; nix develop --command stack --no-nix --system-ghc --no-install-ghc run haga-lambda"