MemAI 0.1.3__tar.gz → 1.0.0__tar.gz

{memai-0.1.3 → memai-1.0.0}/MemAI.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: MemAI
-Version: 0.1.3
+Version: 1.0.0
 Summary: Memory based AI. It doesn't require any parameters and also doesn't require huge amount of data. just needs the sufficient
 Author: Apan Dan
 License-Expression: MIT

{memai-0.1.3 → memai-1.0.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: MemAI
-Version: 0.1.3
+Version: 1.0.0
 Summary: Memory based AI. It doesn't require any parameters and also doesn't require huge amount of data. just needs the sufficient
 Author: Apan Dan
 License-Expression: MIT

{memai-0.1.3 → memai-1.0.0}/memai/Memai.py

@@ -1,11 +1,13 @@
 # The Shaper
 import numpy as np
 
+class DumbnessError(Exception):
+    pass
 
 class Model:
-    def __init__(self, name_id, dataset, modulo=False):
+    def __init__(self, name_id : str, dataset : list, modulo=False):
         self.name_id = name_id
-        self.dataset = dataset
+        self.dataset : list = dataset
         self.modulo  = modulo
 
     def change_dataset(self, new_dataset):
@@ -35,6 +37,8 @@ Estimates the correct output by subtracting the loss from the stored output.
 Calculates the difference between the closest stored input and the given input. Subtracting this difference adjusts the estimated output for the new input."""
 
 
+        
+        
         distances = [
             sum((c - t) ** 2 for c, t in zip(row[0], input))
             for row in self.dataset
@@ -47,6 +51,33 @@ Calculates the difference between the closest stored input and the given input.
         else:
             return c_l[1] - c_l[2] -  sum([a - b for a, b in zip(c_l[0],input)])
 
+
+
+        
+
+    def train(self, input:list, target: int | float):
+        """
+        this function allows the model to learn by letting it try to predict a given input and then putting its try in the dataset.
+        """
+            
+            
+        distances = [
+                    sum((c - t) ** 2 for c, t in zip(row[0], input))
+                    for row in self.dataset
+                ]
+        try:
+            closest_index = distances.index(min(distances))
+        except ValueError or TypeError or Exception:
+            raise DumbnessError("The dataset is empty bro! Do you want the model to be as dumb as 'YOU' bro?")
+        c_l = self.dataset[closest_index]
+        
+        if self.modulo:
+            result = (c_l[1] - c_l[2] -  sum([a - b for a, b in zip(c_l[0],input)])) % self.modulo
+            self.dataset.append([input,result,result-target])
+        else:
+            result =  c_l[1] - c_l[2] -  sum([a - b for a, b in zip(c_l[0],input)]) 
+            self.dataset.append([input,result,result-target])
+
     def get_correct_outputs(self):
             """
             It tries to reverse engineer all the *Expected Outputs* in the datasets by substracting the random output genarated from the loss of it. 

memai-1.0.0/memai/__init__.py

@@ -0,0 +1 @@
+from .Memai import Model

{memai-0.1.3 → memai-1.0.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "MemAI"
-version = "0.1.3"
+version = "1.0.0"
 description = "Memory based AI. It doesn't require any parameters and also doesn't require huge amount of data. just needs the sufficient"
 readme = "README.md"
 authors = [

memai-0.1.3/memai/__init__.py

@@ -1 +0,0 @@
-from .shaper import Model