additional parameter for weight factor for increase or decrease

src/nmreval/models/transitions.py

@@ -4,16 +4,26 @@ from scipy import special as special
 from ..utils import kB
 
 
-class Weight2Phase:
+class Weight:
     type = 'Line shape'
     name = 'Weighting factor'
-    equation = r'A*[0.5 + 0.5 erf[(x-T_{0})/\DeltaT]] + A_{0}'
+    equation = r'A * [0.5 \pm 0.5 erf[(x-T_{0})/\DeltaT]] + A_{0}'
     params = ['T_{0}', r'\DeltaT', 'A', 'A_{0}']
+    choices = [('Direction', 'sign', {'increase': '+', 'decrease': '-'})]
     bounds = [(0, None), (0, None), (None, None), (None, None)]
 
     @staticmethod
-    def func(x, t0, dt, amp, off):
-        return amp*(0.5 + 0.5*special.erf((x-t0)/dt)) + off
+    def func(x: np.ndarray | float, t0: float, dt: float, amp: float, off: float, sign: str = '+') -> np.ndarray | float:
+        if sign not in '+-':
+            raise ValueError(f"`value` is `+` or `-`, not {sign}")
+
+        error_func = 1
+        if sign == '+':
+            error_func += special.erf((x-t0)/dt)
+        else:
+            error_func -= special.erf((x - t0) / dt)
+
+        return amp * error_func / 2. + off
 
 
 class HendricksonBray:
@@ -24,5 +34,5 @@ class HendricksonBray:
     bounds = [(0, None)] * 4
 
     @staticmethod
-    def func(x, a, b, e, w0):
+    def func(x: np.ndarray | float, a: float, b: float, e: float, w0: float) -> np.ndarray | float:
         return a*b / (b + (a-b)*np.exp(-e/kB/x)) + w0