dealing with ceramics in QRF

exposure3_module/exposure.py

@@ -3,7 +3,7 @@ import numpy as np
 from flask import render_template, request
 from functions import SigFigs, Piringer, WilkeChang, SheetRelease, SheetRates, RatePlot
 from functions import Piecewise, PowerLaw
-from qrf_functions import QRF_Apply
+from qrf_functions import QRF_Apply, QRF_Ceramic
 from . import blueprint
 from polymers import Polymers, Polymers3
 from ChemID import ResolveChemical
@@ -135,8 +135,10 @@ def exp_post():
 
     if use_qrf:
         #print('using qrf', file=sys.stderr)
-        ## TODO ceramics
-        diff,domain_extrap = QRF_Apply(density, polytg, smiles, quantiles=[0.03,0.5,0.97])
+        if ceramic:
+            diff,domain_extrap = QRF_Ceramic(density, polytg, quantiles=[0.03,0.5,0.97])
+        else:
+            diff,domain_extrap = QRF_Apply(density, polytg, smiles, quantiles=[0.03,0.5,0.97])
         diff = diff[2] # upper bound
     else:
         ## use categories

qrf_functions.py

@@ -9,6 +9,31 @@ import mordred.descriptors
 import rdkit
 from rdkit import Chem
 
+def QRF_Ceramic(density, polytg, quantiles=[0.03,0.5,0.97]):
+    with open(f'qrf_model_bundle_37.pkl','rb') as f:
+        reg, imp, scaler_X, sub_desc_list = pickle.load(f)
+    df_X = pd.read_excel('qrf_x.xlsx')
+    df_y = pd.read_excel('qrf_y.xlsx')
+    X_all = imp.transform(df_X)
+    X_all_scale = scaler_X.transform(X_all)
+    ## use "worst-case" solute values
+    tmpq = np.array([0.95]*len(sub_desc_list))
+    tmpq[df_X.corrwith(df_y['LogD'])>0] = 0.05 # positive correlations (increase in variable increases D) use low values of variable, negative correlations use high values of variable
+    tmpv = [np.nanquantile(X_all_scale[:,i], q) for i,q in enumerate(tmpq)] # "worst-case" values of scaled descriptors
+    tmps = [polytg if n == 'Polymer_Tg' else (density if n == 'Polymer_Density' else 0) for i,n in enumerate(sub_desc_list)]
+    tmps = scaler_X.transform([tmps])[0] # scaled values of polymer descriptors
+    tmpv = [tmps[i] if n == 'Polymer_Tg' else (tmps[i] if n == 'Polymer_Density' else tmpv[i]) for i,n in enumerate(sub_desc_list)] # merge scaled polymer descriptors with worst-case scaled solute descriptors
+    LogD_pred = reg.predict([tmpv], quantiles=quantiles)
+    D_pred = 10**LogD_pred
+    if len(D_pred.shape)>1:
+        # return 1D array regardless of quantiles setting
+        D_pred = D_pred[0]
+    ## domain extrapolation check
+    dij = QRF_DomainExtrap(reg, X_all_scale, np.array([tmpv]))
+    domain_extrap = dij[0] > 0
+    return D_pred, domain_extrap
+
+
 def QRF_Apply(density, polytg, smiles, quantiles=[0.03,0.5,0.97]):
     with open(f'qrf_model_bundle_37.pkl','rb') as f:
         reg, imp, scaler_X, sub_desc_list = pickle.load(f)
@@ -39,10 +64,10 @@ def QRF_Apply(density, polytg, smiles, quantiles=[0.03,0.5,0.97]):
         D_pred = D_pred[0]
     ## domain extrapolation check
     df_X = pd.read_excel('qrf_x.xlsx')
-    X_train = imp.transform(df_X)
-    X_train_scale = scaler_X.transform(X_train)
-    dij = QRF_DomainExtrap(reg, X_train_scale, descs_scale)
-    domain_extrap = dij > 0
+    X_all = imp.transform(df_X)
+    X_all_scale = scaler_X.transform(X_all)
+    dij = QRF_DomainExtrap(reg, X_all_scale, descs_scale)
+    domain_extrap = dij[0] > 0
     return D_pred, domain_extrap
 
 def QRF_DomainExtrap(estimator, X_train, X_test, D=0.0, return_features=False):

templates/main.html

@@ -62,14 +62,6 @@ This module can be used similarly to the color additive tool, but can be applied
     the total amount of the chemical is known, e.g. from a certificate of analysis.
 </p>
 
-&#x2022; <a href="/exposure3"> Bulk chemicals (excluding color additives) RST (v3) </a>
-
-<p>
-This module can be used similarly to the color additive tool, but can be applied to any non-color additive bulk additive or impurity (surface impurities / manufacturing
-    residuals are excluded) that is present in a polymeric device component provided that the user can justify the chemical is a bulk species and that
-    the total amount of the chemical is known, e.g. from a certificate of analysis. Updated to address additional polymers.
-</p>
-
 &#x2022; <a href="/efficiency"> Extraction efficiency </a>
 
 <p>
@@ -78,8 +70,16 @@ This module can be used to estimate the fraction of the total pool or solvent co
     extraction conditions depending on the endpoint of interest.
 </p>
 
-<!--
-<h2> Modules in development <button type=button class="Info_btn" data-toggle="modal" data-target="#DevModal">&#9432;</button> </h2>-->
+
+<h2> Modules in development <button type=button class="Info_btn" data-toggle="modal" data-target="#DevModal">&#9432;</button> </h2>
+
+&#x2022; <a href="/exposure3"> Bulk chemicals (excluding color additives) (v3) </a>
+
+<p>
+This module can be used similarly to the color additive tool, but can be applied to any non-color additive bulk additive or impurity (surface impurities / manufacturing
+    residuals are excluded) that is present in a polymeric device component provided that the user can justify the chemical is a bulk species and that
+    the total amount of the chemical is known, e.g. from a certificate of analysis. Updated to address additional polymers.
+</p>
 
 <!--<h2> Future modules <button type=button class="Info_btn" data-toggle="modal" data-target="#FutureModal">&#9432;</button> </h2>