X_train = train_set[mylistvariables]
y_train = train_set[myvariablesy]
X_test = test_set[mylistvariables]
y_test = test_set[myvariablesy]
trainedmodels = []
if (docorrelation == 1):
ptmin = 5
ptmax = 7
var_pt = "pt_cand_ML"
var_signal = "signal_ML"
train_set_ptsel = filterdataframe_pt(train_set, var_pt, ptmin, ptmax)
train_set_ptsel_sig, train_set_ptsel_bkg = splitdataframe_sigbkg(
train_set_ptsel, var_signal)
vardistplot(train_set_ptsel_sig, train_set_ptsel_bkg, mylistvariablesall,
"plots")
scatterplot(train_set_ptsel_sig, train_set_ptsel_bkg, mylistvariablesx,
mylistvariablesy, "plots")
correlationmatrix(train_set_ptsel_sig, "plots", "signal")
correlationmatrix(train_set_ptsel_bkg, "plots", "background")
if (doStandard == 1):
X_train = GetDataFrameStandardised(X_train)
if (doPCA == 1):
n_pca = 9
X_train, pca = GetPCADataFrameAndPC(X_train, n_pca)
plotvariancePCA(pca, "plots")
train_set.to_pickle(dataframe + "/dataframetrainsampleN%s.pkl" % (suffix))
test_set.to_pickle(dataframe + "/dataframetestsampleN%s.pkl" % (suffix))
train_set = pd.read_pickle(dataframe + "/dataframetrainsampleN%s.pkl" %
(suffix))
test_set = pd.read_pickle(dataframe + "/dataframetestsampleN%s.pkl" % (suffix))
print("dimension of the dataset", len(train_set))
X_train = train_set[mylistvariables]
y_train = train_set[myvariablesy]
trainedmodels = []
if (docorrelation == 1):
train_set_ptsel_sig, train_set_ptsel_bkg = splitdataframe_sigbkg(
train_set, myvariablesy)
vardistplot(train_set_ptsel_sig, train_set_ptsel_bkg, mylistvariablesall,
plotdir)
scatterplot(train_set_ptsel_sig, train_set_ptsel_bkg, mylistvariablesx,
mylistvariablesy, plotdir)
correlationmatrix(train_set_ptsel_sig, plotdir, "signal")
correlationmatrix(train_set_ptsel_bkg, plotdir, "background")
if (doStandard == 1):
X_train = GetDataFrameStandardised(X_train)
if (doPCA == 1):
n_pca = 5
X_train, pca = GetPCADataFrameAndPC(X_train, n_pca)
plotvariancePCA(pca, plotdir)