Fix the distribution definition in stiffened panel

Model description specifies a dependence between f1 and f2 variables with
Spearman coefficient of -0.8.
Added fields are:
  - the correlation matrix of dimension 10;
  - the Normal copula used for the input variables;
  - the input distribution with the dependence;
  - an independent distribution if needed

Also fix typos in class documentation.

python/doc/usecases/use_case_stiffened_panel.rst

@@ -35,7 +35,7 @@ This use-case implements a simplified model of buckling for a stiffened panel (s
     **Figure 3.** Parameterization of the stiffened panel.
 
 
-This test case is composed of nine random variables:
+This test case is composed of ten random variables:
 
 - :math:`E\sim\mathcal{TN}(\num{110.0e9}, \num{55.0e9}, \num{99.0e9}, \num{121.0e9})` : Young modulus (:math:`\unit{\Pa}`)
 

python/src/usecases/stiffened_panel.py

@@ -63,18 +63,27 @@ class StiffenedPanel:
         Half-width of the stiffener (m) distribution
         `ot.Uniform(0.03762, 0.03838)`
 
+    correlation_matrix : :class:`~openturns.CorrelationMatrix`
+        The correlation matrix used for inputs dependence, mostly identity except for the term representing the interaction between variables :math:`f_1` and :math:`f_2` which is -0.8.
+
+    copula : :class:`~openturns.NormalCopula`
+        The (Normal) copula used to define the distribution of the input parameters.
+
     distribution : :class:`~openturns.JointDistribution`
         The joint distribution of the input parameters.
 
+    independentDistribution : :class:`~openturns.JointDistribution`
+        The joint distribution of the input parameters for the special case of independence.
+
     Examples
     --------
     >>> from openturns.usecases import stiffened_panel
     >>> # Load the stiffened panel model
     >>> panel = stiffened_panel.StiffenedPanel()
     >>> print("Inputs:", panel.model.getInputDescription())
-    Inputs: [F,L,a,De,di,E]
+    Inputs: [E,nu,h_c,ell,f_1,f_2,t,a,b_0,p]
     >>> print("Outputs:", panel.model.getOutputDescription())
-    [Deflection,Left angle,Right angle]
+    Outputs: [(N_{xy})_{cr}]
     """
 
     def __init__(self):
@@ -135,17 +144,30 @@ def __init__(self):
         self.p = ot.Uniform(0.03762, 0.03838)
         self.p.setDescription(["p (m)"])
 
-        self.distribution = ot.JointDistribution(
-            [
-                self.youngModulus,
-                self.nu,
-                self.h_c,
-                self.ell,
-                self.f_1,
-                self.f_2,
-                self.t,
-                self.a,
-                self.b_0,
-                self.p,
-            ]
+        # correlation matrix
+        self.correlation_matrix = ot.CorrelationMatrix(self.dim)
+        self.correlation_matrix[4, 5] = -0.8
+        self.copula = ot.NormalCopula(
+            ot.NormalCopula.GetCorrelationFromSpearmanCorrelation(
+                self.correlation_matrix
+            )
         )
+
+        varList = [
+            self.youngModulus,
+            self.nu,
+            self.h_c,
+            self.ell,
+            self.f_1,
+            self.f_2,
+            self.t,
+            self.a,
+            self.b_0,
+            self.p,
+        ]
+
+        # distribution
+        self.distribution = ot.JointDistribution(varList, self.copula)
+
+        # special case: independent distribution
+        self.independentDistribution = ot.JointDistribution(varList)