Update experiments

data/experiments/ch_activation_experiments_a.csv

@@ -0,0 +1,25 @@
+,,Conditions,,,,,,Results,,,,,Weighing,,,,,,,,,
+Data file,Type,Solvent,Ligand,ResT /min,Temp /°C,Mol%,Yield /%,SM area,Product area,Biphenyl area,SM conc,Product conc,Volume /mL,Conc SM /M,Mol% catalyst,SM /g,Pd(OAc)2 /g,Ligand Mr,Ligand /g,NEt3 /mL,Biphenyl /g,Biphenyl /M
+TrainingData001,Training,DMSO (4),DPEPhos (4),60,140,10,46.09,,3591526,6132457,0,0.046087689,5,0.1,10,0.1008,0.0112,538.5,0.05385,0.104545455,0.031,0.04025974
+TrainingData002,Training,DMSO (4),JohnPhos (1),45,70,3,1.31,2704169,260938,13341583,0.060887904,0.001305763,5,0.1,3,0.1008,0.00336,298.4,0.008952,0.104545455,0.0263,0.034155844
+TrainingData003,Training,Toluene (1),DPEPhos (4),42,61,8,0,,,,#DIV/0!,#DIV/0!,5,0.1,8,0.1008,0.00896,538.5,0.04308,0.104545455,0.0245,0.031818182
+TrainingData004,Training,DMA (2),JohnPhos (1),9,135,5,37.91,1333228,4837572,8941341,0.047006689,0.037906406,5,0.1,5,0.1008,0.0056,298.4,0.01492,0.104545455,0.0276,0.035844156
+TrainingData005,Training,MeCN (3),JohnPhos (1),11,134,2,16.87,1614954,2408002,9600861,0.050914872,0.016872183,5,0.1,2,0.1008,0.00224,298.4,0.005968,0.104545455,0.0265,0.034415584
+TrainingData006,Training,MeCN (3),JohnPhos (1),19,80,5,69.7,723374,16037972,15186143,0.014146127,0.069703414,5,0.1,5,0.1008,0.0056,298.4,0.01492,0.104545455,0.026,0.033766234
+TrainingData007,Training,NMP (5),JohnPhos (1),47,143,4,40.92,,7980309,14951125,0,0.040919613,5,0.1,4,0.1008,0.00448,298.4,0.011936,0.104545455,0.0302,0.039220779
+TrainingData008,Training,DMSO (4),SPhos (2),29,122,4,0,1001820,0,11654167,0.02651069,0,5,0.1,4,0.1008,0.00448,410.5,0.01642,0.104545455,0.027,0.035064935
+TrainingData009,Training,NMP (5),Xphos (3),52,112,8,55.69,647744,10707780,13665700,0.015159272,0.055693476,5,0.1,8,0.1008,0.00896,476.7,0.038136,0.104545455,0.028,0.036363636
+,Training,NMP (5),DPEPhos (4),26,53,6,0,,,,#DIV/0!,#DIV/0!,5,0.1,6,0.1008,0.00672,,0,0.104545455,,0
+TrainingData011,Training,Toluene (1),Xphos (3),6,83,3,3.01,3037966,565891,15533650,0.072600826,0.003005532,5,0.1,3,0.1008,0.00336,476.7,0.014301,0.104545455,0.0325,0.042207792
+TrainingData012,Training,MeCN (3),Xphos (3),39,90,9,6.33,2462821,1180830,12983910,0.059364488,0.006325738,5,0.1,9,0.1008,0.01008,476.7,0.042903,0.104545455,0.0274,0.035584416
+TrainingData013,Training,DMA (2),Xphos (3),51,104,9,16.11,1205010,3186820,12801630,,0.01611429,5,0.1,9,0.1008,0.01008,476.7,0.042903,0.104545455,0.0255,0.033116883
+TrainingData014,Training,DMA (2),DPEPhos (4),22,67,4,4.04,2696188,825115,12949428,,0.004043735,5,0.1,4,0.1008,0.00448,538.5,0.02154,0.104545455,0.025,0.032467532
+TrainingData015,Training,DMSO (4),Xphos (3),33,115,7,0,,,,,#DIV/0!,5,0.1,7,0.1008,0.00784,476.7,0.033369,0.104545455,0.0258,0.033506494
+,Training,Toluene (1),SPhos (2),19,57,8,0,,,,,#DIV/0!,5,0.1,8,0.1008,0.00896,,0,0.104545455,,0
+,Optimization,NMP (5),Xphos (3),60,102,10,56.2,,,,,,,,,,,,,,,
+,Optimization,MeCN (3),JohnPhos (1),20,69,6,5.9,,,,,,,,,,,,,,,
+,Optimization,Toluene (1),JohnPhos (1),53,141,8,26.4,,,,,,,,,,,,,,,
+Bumped up to 100 restarts and 2000 raw samples,Optimization,NMP (5),Xphos (3),60,123,9,42.4,,,,,,,,,,,,,,,
+,Optimization,NMP (5),Xphos (3),53,89,9,74.7,,,,,,,,,,,,,,,
+,Optimization,NMP (5),Xphos (3),46,88,9,46,,,,,,,,,,,,,,,
+,Optimization,NMP (5),Xphos (3),60,85,8,72.4,,,,,,,,,,,,,,,

data/experiments/ch_activation_experiments_b.csv

@@ -0,0 +1,4 @@
+,,Conditions,,,,,,Results,,,,,Weighing,,,,,,,,,
+Data file,Type,Solvent,Ligand,ResT /min,Temp /°C,Mol%,Yield /%,SM area,Product area,Biphenyl area,SM conc,Product conc,Volume /mL,Conc SM /M,Mol% catalyst,SM /g,Pd(OAc)2 /g,Ligand Mr,Ligand /g,NEt3 /mL,Biphenyl /g,Biphenyl /M
+,Optimization,NMP,Xphos,19,75,4,17.4,,,,,,,,,,,,,,,
+Changed to qNEI,Optimization,Toluene,DPEPhos,50,121,8,21.06,,,,,,,,,,,,,,,

data/kinetic_models_old/.gitignore

@@ -0,0 +1 @@
+/grid_artifacts

data/kinetic_models_old/download_grid.sh

@@ -0,0 +1,5 @@
+#!/bin/bash
+ARGS=("$@")
+grid artifacts "${ARGS[@]}"
+rm -r data/kinetic_models/grid_artifacts/*/**/dist
+rm -r data/kinetic_models/grid_artifacts/*/**/multitask.egg-info

data/kinetic_models_old/grid_artifacts.dvc

@@ -0,0 +1,5 @@
+outs:
+- md5: 2967669ca15cc6f2c5466216b749e8f8.dir
+  size: 8887238
+  nfiles: 1444
+  path: grid_artifacts

dvc.lock

@@ -12,7 +12,7 @@ stages:
       size: 17487
     outs:
     - path: data/baumgartner_suzuki/ord/
-      md5: 703e932a228102171974236b0d8fc701.dir
+      md5: b166c964caa6627b76e6b628962decaa.dir
       size: 446426
       nfiles: 2
   suzuki_benchmark_training:

multitask/etl_baumgartner_C-N.py

@@ -71,6 +71,8 @@ def main(
     reactions.extend(
         df_rxn_data.progress_apply(inner_loop, axis=1, args=(df_stock_solutions,))
     )
+    for i, reaction in enumerate(reactions):
+        reaction.reaction_id = df_rxn.iloc[i]["Overall "]
 
     # Create dataset
     dataset = Dataset()

multitask/mt.py

@@ -167,8 +167,8 @@ def suggest_experiments(self, num_experiments, prev_res: DataSet = None, **kwarg
         inputs, output = self.transform.transform_inputs_outputs(
             data,
             categorical_method=self.categorical_method,
-            standardize_inputs=True,
-            standardize_outputs=True,
+            min_max_scale_inputs=True,
+            min_max_scale_outputs=True,
         )
 
         # Categorial transformation
@@ -184,7 +184,7 @@ def suggest_experiments(self, num_experiments, prev_res: DataSet = None, **kwarg
 
         # Add column to inputs indicating task
         task_data = data["task"].dropna().to_numpy()
-        if data.shape[0] != data.shape[0]:
+        if task_data.shape[0] != data.shape[0]:
             raise ValueError("Pretraining data must have a task for every row.")
         task_data = np.atleast_2d(task_data).T
         inputs_task = np.append(
@@ -195,7 +195,7 @@ def suggest_experiments(self, num_experiments, prev_res: DataSet = None, **kwarg
         objective = self.domain.output_variables[0]
         if not objective.maximize:
             output = -1.0 * output
-        fbest_scaled = output.max()
+        fbest_scaled = output[objective.name].max()
 
         # Train model
         if self.brute_force_categorical and self.categorical_method is None:
@@ -245,10 +245,10 @@ def suggest_experiments(self, num_experiments, prev_res: DataSet = None, **kwarg
             results, _ = optimize_acqf_mixed(
                 acq_function=self.acq,
                 bounds=self._get_bounds(),
-                num_restarts=kwargs.get("num_restarts", 5),
+                num_restarts=kwargs.get("num_restarts", 100),
                 fixed_features_list=fixed_features_list,
                 q=num_experiments,
-                raw_samples=kwargs.get("raw_samples", 100),
+                raw_samples=kwargs.get("raw_samples", 2000),
             )
         else:
             if self.acquistion_function == "EI":
@@ -268,9 +268,9 @@ def suggest_experiments(self, num_experiments, prev_res: DataSet = None, **kwarg
             results, _ = optimize_acqf(
                 acq_function=self.acq,
                 bounds=self._get_bounds(),
-                num_restarts=kwargs.get("num_restarts", 20),
+                num_restarts=kwargs.get("num_restarts", 100),
                 q=num_experiments,
-                raw_samples=kwargs.get("raw_samples", 100),
+                raw_samples=kwargs.get("raw_samples", 2000),
             )
 
         # Convert result to datset
@@ -287,7 +287,9 @@ def suggest_experiments(self, num_experiments, prev_res: DataSet = None, **kwarg
                     result[v.name] = result[v.name].replace(cat_mapping)
 
         result = self.transform.un_transform(
-            result, categorical_method=self.categorical_method, standardize_inputs=True
+            result,
+            categorical_method=self.categorical_method,
+            min_max_scale_inputs=True,
         )
 
         # Add metadata
@@ -634,10 +636,10 @@ def suggest_experiments(self, num_experiments, prev_res: DataSet = None, **kwarg
             results, _ = optimize_acqf_mixed(
                 acq_function=self.acq,
                 bounds=self._get_bounds(),
-                num_restarts=kwargs.get("num_restarts", 5),
+                num_restarts=kwargs.get("num_restarts", 100),
                 fixed_features_list=fixed_features_list,
                 q=num_experiments,
-                raw_samples=kwargs.get("raw_samples", 100),
+                raw_samples=kwargs.get("raw_samples", 2000),
             )
         else:
             if self.acquistion_function == "EI":
@@ -659,9 +661,9 @@ def suggest_experiments(self, num_experiments, prev_res: DataSet = None, **kwarg
             results, _ = optimize_acqf(
                 acq_function=self.acq,
                 bounds=self._get_bounds(),
-                num_restarts=kwargs.get("num_restarts", 5),
+                num_restarts=kwargs.get("num_restarts", 100),
                 q=num_experiments,
-                raw_samples=kwargs.get("raw_samples", 100),
+                raw_samples=kwargs.get("raw_samples", 2000),
             )
 
         # Convert result to datset

nbs/baumgartner_C-N v3.ipynb

@@ -2750,7 +2750,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.7.4"
+   "version": "3.9.10"
   },
   "toc": {
    "base_numbering": 1,

nbs/experiments.ipynb → nbs/bh_experiments.ipynb

@@ -1169,7 +1169,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.9.10"
+   "version": "3.9.12"
   }
  },
  "nbformat": 4,