Added Batch example from Ind. Eng. Chem. Res. 2003, 42, 15, 3592–3601

examples/batch_dryer_case_study.py

@@ -0,0 +1,78 @@
+# -*- coding: utf-8 -*-
+"""
+Created on Tue Apr 26 22:33:38 2022
+
+@author: salva
+"""
+
+
+import pandas as pd
+import numpy as np
+import pyphi_batch as phibatch
+import matplotlib.pyplot as plt
+import pyphi_plots as pp
+
+
+bdata        = pd.read_excel('Batch Dryer Case Study.xlsx',sheet_name='Trajectories')
+cqa          = pd.read_excel('Batch Dryer Case Study.xlsx',sheet_name='ProductQuality')
+char         = pd.read_excel('Batch Dryer Case Study.xlsx',sheet_name='ProcessCharacteristics')
+initial_chem = pd.read_excel('Batch Dryer Case Study.xlsx',sheet_name='InitialChemistry')
+cat          = pd.read_excel('Batch Dryer Case Study.xlsx',sheet_name='classifiers')
+
+phibatch.plot_var_all_batches(bdata)
+#%%
+samples={'Deagglomerate':20,'Heat':30,'Cooldown':40};
+bdata_aligned_phase=phibatch.phase_simple_align(bdata,samples)
+phibatch.plot_var_all_batches(bdata_aligned_phase,
+                  plot_title='simple alignment /phase',
+                  phase_samples=samples)
+#%%
+mpca_obj=phibatch.mpca(bdata_aligned_phase,3,phase_samples=samples)
+mpca_mon=phibatch.monitor(mpca_obj, bdata_aligned_phase)
+mon_data =phibatch.monitor(mpca_obj, bdata_aligned_phase,which_batch='Batch 5')
+
+#Show forecast of x
+
+var='Dryer Temp'
+point_in_time=20
+batch2forecast='Batch 5'
+forecast=mon_data[0]['forecast']
+mdata=bdata_aligned_phase[bdata_aligned_phase['Batch number']==batch2forecast]
+
+plt.figure()
+f=forecast[point_in_time]
+plt.plot(mdata[var][:point_in_time],'o',label='Measured')
+aux=[np.nan]*point_in_time
+aux.extend(f[var].values[point_in_time:].tolist())
+plt.plot(np.array(aux),label='Forecast')
+plt.plot(mdata[var][point_in_time:],'o',label='Known trajectory',alpha=0.3)
+plt.xlabel('sample')
+plt.ylabel(var)
+plt.legend()
+plt.title('Forecast for '+var+' at sample '+str(point_in_time)+' for '+batch2forecast )
+
+#%%
+mpls_obj=phibatch.mpls(bdata_aligned_phase,cqa,3,phase_samples=samples)
+pp.r2pv(mpls_obj)
+pp.score_scatter(mpls_obj, [1,2],CLASSID=cat,colorby='Quality')
+mpls_obj=phibatch.monitor(mpls_obj, bdata_aligned_phase)
+mon_data =phibatch.monitor(mpls_obj, bdata_aligned_phase,which_batch='Batch 5')
+
+
+#%%
+mpls_obj=phibatch.mpls(bdata_aligned_phase,cqa,5,zinit=initial_chem,
+                       phase_samples=samples)
+phibatch.loadings(mpls_obj,1)
+pp.r2pv(mpls_obj)
+mpls_obj=phibatch.monitor(mpls_obj, bdata_aligned_phase,zinit=initial_chem)
+mon_data =phibatch.monitor(mpls_obj, bdata_aligned_phase,which_batch='Batch 5',zinit=initial_chem)
+
+#%%
+
+mpls_obj=phibatch.mpls(bdata_aligned_phase,cqa,5,zinit=initial_chem,
+                      phase_samples=samples,mb_each_var='True',cross_val=5)
+phibatch.loadings(mpls_obj,1)
+pp.r2pv(mpls_obj)
+pp.score_scatter(mpls_obj, [1,2],CLASSID=cat,colorby='Quality')
+mpls_obj=phibatch.monitor(mpls_obj, bdata_aligned_phase,zinit=initial_chem)
+mon_data =phibatch.monitor(mpls_obj, bdata_aligned_phase,which_batch='Batch 5',zinit=initial_chem)