Skip to content

Commit

Permalink
imporvements
Browse files Browse the repository at this point in the history
  • Loading branch information
@lordmikerahl
lordmikerahl committed Mar 27, 2024
1 parent f50c2a3 commit 5241c6f
Show file tree
Hide file tree
Showing 5 changed files with 56 additions and 53 deletions.
109 changes: 56 additions & 53 deletions baybe_hack.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -108,16 +108,16 @@
"NumericalDiscreteParameter(\n",
" name=\"Time (h)\",\n",
" values=np.arange(1, 25, 1)\n",
" # tolerance = 0.004\n",
" # tolerance = 0.004, assume certain experimental noise for each parameter measurement?\n",
"),\n",
"NumericalDiscreteParameter(\n",
" name=\"pH\",\n",
" values=np.arange(-1, 15.1, 0.1)\n",
" # tolerance = 0.004\n",
" ), \n",
"NumericalContinuousParameter( # Set this as continuous, the values seem quite small?\n",
"NumericalDiscreteParameter( # Set this as continuous, the values seem quite small?\n",
" name=\"Inhibitor Concentration (M)\",\n",
" bounds=(0, 0.02)\n",
" values=np.arange(0, 0.1, 0.01), # Remove data outliers like 0.1?\n",
" # tolerance = 0.004\n",
" ),\n",
"NumericalDiscreteParameter(\n",
Expand Down Expand Up @@ -161,33 +161,21 @@
"SubstanceParameter(\n",
" name=\"SMILES\",\n",
" data={\n",
" # df_AA2024 SMILES column\n",
" # INCORPORATE TRAINING DATA FROM DATAFRAME SMILES COLUMN\n",
" \"Water\": \"O\",\n",
" \"1-Octanol\": \"CCCCCCCCO\",\n",
" \"Toluene\": \"CC1=CC=CC=C1\",\n",
" },\n",
" encoding=\"MORDRED\", # optional\n",
" decorrelate=0.7, # optional\n",
" encoding=\"MORDRED\", # Can be also RDKIT or MORGAN_FP - WHICH IS BETTER?\n",
" decorrelate=0.7, # Change threshold to avoid overfitting?\n",
")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"The encoding option defines what kind of descriptors are calculated:\n",
"\n",
"MORDRED: 2D descriptors from the Mordred package. Since the original package is now unmaintained, baybe requires the community replacement mordredcommunity\n",
"\n",
"RDKIT: 2D descriptors from the RDKit package\n",
"\n",
"MORGAN_FP: Morgan fingerprints calculated with RDKit (1024 bits, radius 4)\n",
"\n",
"These calculations will typically result in 500 to 1500 numbers per molecule. **To avoid detrimental effects on the surrogate model fit, we reduce the number of descriptors via decorrelation before using them.** For instance, the decorrelate option in the example above specifies that only descriptors with a correlation lower than 0.7 to any other descriptor will be kept. This usually reduces the number of descriptors to 10-50, depending on the specific items in data.\n",
"\n",
"**WARNING:**\n",
"The descriptors calculated for a SubstanceParameter were developed to describe small molecules and are not suitable for other substances. If you deal with large molecules like polymers or arbitrary substance mixtures, we recommend to provide your own descriptors via the CustomParameter."
"These calculations will typically result in 500 to 1500 numbers per molecule. **To avoid detrimental effects on the surrogate model fit, we reduce the number of descriptors via decorrelation before using them.** For instance, the decorrelate option in the example above specifies that only descriptors with a correlation lower than 0.7 to any other descriptor will be kept. This usually reduces the number of descriptors to 10-50, depending on the specific items in data."
]
},
{
Expand All @@ -201,23 +189,9 @@
},
{
"cell_type": "code",
"execution_count": 5,
"execution_count": null,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"CustomDiscreteParameter(name='Polymer', data= Glass_Transition_TempC Weight_kDalton\n",
"Polymer A 20 120\n",
"Polymer B -71 32\n",
"Polymer C -39 241, decorrelate=True, encoding=<CustomEncoding.CUSTOM: 'CUSTOM'>)"
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"outputs": [],
"source": [
"\"\"\"\n",
"from baybe.parameters import CustomDiscreteParameter\n",
Expand Down Expand Up @@ -353,9 +327,16 @@
"campaign = Campaign(searchspace, objective, strategy)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Get recommendations"
]
},
{
"cell_type": "code",
"execution_count": 18,
"execution_count": 21,
"metadata": {},
"outputs": [
{
Expand All @@ -365,23 +346,21 @@
"\n",
"\n",
"Recommended experiments: \n",
"| | Time (h) | pH | Salt Concentration (M) | Inhibitor Concentration (M) |\n",
"|-------:|-----------:|-----:|-------------------------:|------------------------------:|\n",
"| 492324 | 16 | 2.4 | 0.75 | 0.0150329 |\n",
"| 341299 | 11 | 7.8 | 0.01 | 0.00234041 |\n",
"| 470340 | 15 | 7.6 | 0 | 0.00838565 |\n"
"| | Time (h) | pH | Salt Concentration (M) | Inhibitor Concentration (M) |\n",
"|------:|-----------:|-----:|-------------------------:|------------------------------:|\n",
"| 11808 | 1 | 4.8 | 1.5 | 0.00858356 |\n"
]
}
],
"source": [
"new_rec = campaign.recommend(batch_size=3)\n",
"new_rec = campaign.recommend(batch_size=1) # TEST with different batch sizes for optimal performance\n",
"print(\"\\n\\nRecommended experiments: \")\n",
"print(new_rec.to_markdown())"
]
},
{
"cell_type": "code",
"execution_count": 19,
"execution_count": 22,
"metadata": {},
"outputs": [
{
Expand All @@ -391,27 +370,26 @@
"\n",
"\n",
"Recommended experiments with measured values: \n",
"| | Time (h) | pH | Salt Concentration (M) | Inhibitor Concentration (M) | efficiency |\n",
"|-------:|-----------:|-----:|-------------------------:|------------------------------:|-------------:|\n",
"| 492324 | 16 | 2.4 | 0.75 | 0.0150329 | 0.1 |\n",
"| 341299 | 11 | 7.8 | 0.01 | 0.00234041 | 0.2 |\n",
"| 470340 | 15 | 7.6 | 0 | 0.00838565 | 0.3 |\n"
"| | Time (h) | pH | Salt Concentration (M) | Inhibitor Concentration (M) |\n",
"|------:|-----------:|-----:|-------------------------:|------------------------------:|\n",
"| 11808 | 1 | 4.8 | 1.5 | 0.00858356 |\n"
]
}
],
"source": [
"new_rec[\"efficiency\"] = [0.1,0.2,0.3]\n",
"# Get and input efficiency value from Excel table, for specific SMILES component first, \n",
"# then for the closest values of the rest of the parameters\n",
"\n",
"new_rec[\"efficiency\"] = [0.1]\n",
"print(\"\\n\\nRecommended experiments with measured values: \")\n",
"print(new_rec.to_markdown())"
]
},
{
"cell_type": "code",
"execution_count": null,
"cell_type": "markdown",
"metadata": {},
"outputs": [],
"source": [
"# Merge all results into a dataframe"
"### Merge all results into a dataframe"
]
},
{
Expand All @@ -424,6 +402,31 @@
"print(\"\\n\\nAll experiments with measured values: \")\n",
"print(results.to_markdown())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Transfer learning + Initial Data INFO"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"https://emdgroup.github.io/baybe/examples/Transfer_Learning/basic_transfer_learning.html\n",
"\n",
"https://emdgroup.github.io/baybe/userguide/transfer_learning.html"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"https://emdgroup.github.io/baybe/examples/Backtesting/full_initial_data.html\n",
"\n",
"https://emdgroup.github.io/baybe/examples/Backtesting/full_lookup.html"
]
}
],
"metadata": {
Expand Down
Binary file added data/filtered_AA1000.xlsx
View file
Binary file not shown.
Binary file modified data/filtered_AA2024.xlsx
View file
Binary file not shown.
Binary file modified data/filtered_Al.xlsx
View file
Binary file not shown.
Binary file removed data/filtered_full.xlsx
View file
Binary file not shown.

0 comments on commit 5241c6f

Please sign in to comment.