use proper numpy eulers constant, better irt model abstraction

app/lib/irt/models/base.py

@@ -0,0 +1,14 @@
+import numpy as np
+
+class Base:
+
+  def __init__(self, model_params, kwargs):
+    self.model_params = model_params
+    # check if exists, if not error out
+    self.b_param = kwargs['b_param']
+    self.e = np.exp
+    self.theta = kwargs['theta']
+
+  @classmethod
+  def ability_estimate(self) -> float:
+    return 0.0

app/lib/irt/models/rasch.py

@@ -1,14 +1,8 @@
-import numpy as np
-
 from girth import ability_mle
 
-class Rasch:
+from lib.irt.models.base import *
 
-  def __init__(self, model_params, kwargs):
+class Rasch(Base):
-    self.model_params = model_params
-    self.b_param = kwargs['b_param']
-    self.e = 2.71828
-    self.theta = kwargs['theta']
 
   def result(self):
     return 0.0

app/lib/irt/models/three_parameter_logistic.py

@@ -1,11 +1,6 @@
-class ThreeParameterLogistic:
+from lib.irt.models.base import *
 
-    def __init__(self, model_params, kwargs):
+class ThreeParameterLogistic(Base):
-        self.model_params = model_params
-        # check if exists, if not error out
-        self.b_param = kwargs['b_param']
-        self.e = 2.71828
-        self.theta = kwargs['theta']
 
     # contains the primary 3pl function, determining the probably of an inidividual
     # that an individual at a certain theta would get a particular question correct
@@ -14,9 +9,5 @@ class ThreeParameterLogistic:
     def result(self):
         a = self.model_params.a_param
         c = self.model_params.c_param
-        return c + (1 - c) * (1 / (1 + self.e**(-a *
+        return c + (1 - c) * (1 / (1 + self.e(-a *
                                                 (self.theta - self.b_param))))
-
-    @classmethod
-    def ability_estimate(self) -> float:
-        return 0.0