diff --git a/src/Class_Mod/LWPLSR_.py b/src/Class_Mod/LWPLSR_.py
index e5682ac39c37e85dd6871faf0e1646b5c5626e30..5e86c902e4926d77b818531284492abffa9d584c 100644
--- a/src/Class_Mod/LWPLSR_.py
+++ b/src/Class_Mod/LWPLSR_.py
@@ -1,19 +1,44 @@
-import pandas as pd
-import streamlit
from Packages import *
from Class_Mod.Miscellaneous import *
class LWPLSR:
"""
- The UMAP dimension reduction algorithm from scikit learn
+ The lwpls regression model from Jchemo (M. Lesnoff)
"""
- def __init__(self, x_train, x_test, y_train, y_test):
+ def __init__(self, x_train, y_train, x_test, y_test):
"""Initiate the LWPLSR and prepare data for Julia computing."""
- self.x_train, self.y_train, self.x_test, self.y_test = x_train, x_test, y_train, y_test
+ self.x_train, self.y_train, self.x_test, self.y_test = x_train, y_train, x_test, y_test
# prepare to send dataframes to julia and Jchemo
- jl.x_train,jl.y_train,jl.x_test,jl.y_test = self.x_train, self.y_train, self.x_test, self.y_test
+ jl.x_train, jl.y_train, jl.x_test, jl.y_test = self.x_train, self.y_train, self.x_test, self.y_test
+ # Pre-treatment of x_train and x_test
+ jl.seval("""
+ # using DataFrames
+ # using Pandas
+ using Jchemo
+ mod1 = Jchemo.model(snv; centr = true, scal = true)
+ mod2 = Jchemo.model(savgol; npoint = 15, deriv = 1, degree = 2)
+ mod = Jchemo.pip(mod1, mod2)
+ Jchemo.fit!(mod, x_train)
+ x_train = Jchemo.transf(mod1, x_train)
+ Jchemo.fit!(mod, x_test)
+ x_test = Jchemo.transf(mod1, x_test)
+ """)
+ jl.seval("""
+ ntrain = nro(x_train)
+ segm = segmkf(ntrain, 4; rep = 5)
+ nlvdis = [5; 10; 15] ; metric = [:mah]
+ h = [1; 2; 6; Inf] ; k = [10; 30; 100]
+ nlv = 0:15
+ pars = mpar(nlvdis = nlvdis, metric = metric, h = h, k = k)
+ println(pars)
+ """)
+
+ # initialize vars from the class
+ y_shape = y_test.shape
self.scores = pd.DataFrame
self.predicted_results_on_test = pd.DataFrame
+ self.pred = np.zeros(shape=(y_shape[0], 1))
+ self.mod = ""
def Jchemo_lwplsr(self):
"""Send data to Julia to compute lwplsr.
@@ -29,38 +54,70 @@ class LWPLSR:
self.predicted_results_on_test (DataFrame):
"""
# launch Julia Jchemo lwplsr
- jl.eval("""
- using Pandas
+ jl.seval("""
using DataFrames
+ using Pandas
using Jchemo
- nlvdis = 5 ; metric = :mah
- h = 1 ; k = 200 ; nlv = 15 #; scal = true
- mod = Jchemo.model(Jchemo.lwplsr; nlvdis, metric, h, k, nlv)
- Jchemo.fit!(mod, X_train, y_train)
- # Jchemo.pnames(mod)
- # Jchemo.pnames(mod.fm)
- # predictions on test data calculation
- res = Jchemo.predict(mod, X_test)
- # Jchemo.pnames(res)
+ x_train |> Pandas.DataFrame |> DataFrames.DataFrame
+ y_train |> Pandas.DataFrame |> DataFrames.DataFrame
+ x_test |> Pandas.DataFrame |> DataFrames.DataFrame
+ y_test |> Pandas.DataFrame |> DataFrames.DataFrame
""")
- resjp = jl.seval("""
- Pandas.DataFrame(res.pred)
+ # Create LWPLSR model and fit
+ # jl.seval("""
+ # nlvdis = 5 ; metric = :mah
+ # h = 1 ; k = 200 ; nlv = 15 #; scal = true
+ # mod = Jchemo.model(Jchemo.lwplsr; nlvdis, metric, h, k, nlv)
+ # # Fit model
+ # Jchemo.fit!(mod, x_train, y_train)
+ # """)
+
+ jl.seval("""
+ mod = Jchemo.model(Jchemo.lwplsr)
+ res = Jchemo.gridcv(mod, x_train, y_train; segm, score = Jchemo.rmsep, pars, nlv, verbose = true).res
+ # u = findall(res.y1 .== minimum(res.y1))[1]
+ # mod = Jchemo.model(lwplsr; nlvdis = res.nlvdis[u], metric = res.metric[u], h = res.h[u], k = res.k[u], nlv = res.nlv[u]) ;
+ # Jchemo.fit!(mod, x_train, y_train)
""")
- scoresjp = jl.seval("""
- Jchemo.mse(res.pred, y_test)
+ self.mod = jl.mod
+
+ def Jchemo_lwplsr_predict(self):
+ # Predictions on x_test and store in self.pred
+ self.pred = jl.seval("""
+ res = Jchemo.predict(mod, x_test)
+ res.pred
""")
- self.scores = pd.Dataframe(scoresjp)
- self.predicted_results_on_test = pd.Dataframe(resjp)
+ # convert predicted data from x_test to Pandas DataFrame
+ self.predicted_results_on_test = pd.DataFrame(self.pred)
@property
def pred_data_(self):
return self.predicted_results_on_test, self.predicted_results_on_test, self.predicted_results_on_test
- # @property
- # def model_(self):
- # return self.trained
+ @property
+ def model_(self):
+ return self.mod
@property
def metrics_(self):
- # self.scores = pd.DataFrame(self.scores, index=['test'])
- return self.scores
\ No newline at end of file
+ jl.pred = self.pred
+ st.dataframe(self.pred)
+ st.dataframe(self.predicted_results_on_test)
+ st.write('starting metrics')
+ jl.seval("""
+ using Jchemo
+ """)
+ scorermsep = jl.seval("""
+ first(Jchemo.rmsep(pred, y_test))
+ """)
+ scoremr2 = jl.seval("""
+ first(Jchemo.r2(pred, y_test))
+ """)
+ scorerpd = jl.seval("""
+ first(Jchemo.rpd(pred, y_test))
+ """)
+ scoremsep = jl.seval("""
+ first(Jchemo.sep(pred, y_test))
+ """)
+ self.scores = pd.DataFrame([[scoremr2, scorermsep, scoremsep, scorerpd]], columns=['r2', 'rmsep', 'msep', 'rpd'], index=['scores'])
+ return self.scores
diff --git a/src/pages/2-model_creation.py b/src/pages/2-model_creation.py
index cfed211cd23bdeb57c997ec2eb27fa521bfdcd57..986d249bb2c254db09665c5e6137f7d70b88824b 100644
--- a/src/pages/2-model_creation.py
+++ b/src/pages/2-model_creation.py
@@ -134,9 +134,13 @@ if not spectra.empty and not y.empty:
reg_model = Reg.model_
#M2.dataframe(Pin.pred_data_)
elif regression_algo == reg_algo[2]:
- # Reg = LWPLSR(x_train = X_train, x_test = X_test, y_train = y_train, y_test = y_test)
- # LWPLSR.Jchemo_lwplsr(Reg)
- pass
+ x_train, y_train, x_test, y_test = X_train.to_numpy(), y_train.to_numpy(), X_test.to_numpy(), y_test.to_numpy()
+ Reg = LWPLSR(x_train, y_train, x_test, y_test)
+ LWPLSR.Jchemo_lwplsr(Reg)
+ reg_model = Reg.model_
+ # LWPLSR.Jchemo_lwplsr_predict(Reg)
+ # st.dataframe(Reg.pred_data_)
+ # st.dataframe(Reg.metrics_)
elif regression_algo == reg_algo[3]:
s = M1.number_input(label='Enter the maximum number of intervals', min_value=1, max_value=6, value=3)