added solve_triangular to frontend

ivy/functional/frontends/torch/linalg.py

@@ -273,6 +273,51 @@ def solve(A, B, *, left=True, out=None):
     return ivy.solve(A, B, out=out)
 
 
+@to_ivy_arrays_and_back
+@with_supported_dtypes(
+    {"2.0.1 and below": ("float32", "float64", "complex32", "complex64")}, "torch"
+)
+def solve_triangular(A, B, *, upper, left=True, unitriangular=False, out=None):
+    # TOD0: Implement left and out
+    X = ivy.zeros_like(B)
+
+    #check if A is triangular
+    if ivy.all(ivy.triu(A)) != 0 or ivy.all(ivy.tril(A)) != 0:
+        raise RuntimeError("Matrix is not triangular")
+
+    #check if matrix is inversible
+    for row_idx, _ in enumerate(A):
+        if A[row_idx][row_idx] == 0:
+            raise RuntimeError("Matrix is not invertible")
+
+    #need to make leading variables 1
+    if unitriangular == False:
+        if upper == True:
+            for row_idx, row in enumerate(A):
+                div = A[row_idx][row_idx]
+                row[:] = [x / div for x in row]
+                B[row_idx][:] = [y / div for y in B[row_idx]]
+        else:
+            for row_idx, row in enumerate(A):
+                div = A[row_idx][0]
+                row[:] = [x / div for x in row]
+                B[row_idx][:] = [y / div for y in B[row_idx]]
+
+    #equation for lower
+    if upper == False:
+        for i in range(len(A)):
+            X[i] = (B[i] - ivy.dot(A[i][:i], X[:i])) / A[i][i]
+
+        return X
+
+    #equation for upper
+    if upper == True:
+        for i in range(len(A) - 1, -1, -1):
+            X[i] = (B[i] - ivy.dot(A[i][i:], X[i:])) / A[i][i]
+
+        return X
+
+
 @to_ivy_arrays_and_back
 @with_supported_dtypes(
     {"2.0.1 and below": ("float32", "float64", "complex32", "complex64")}, "torch"