An attempt to build a simple partial-ish differentiator using automatic-ish differentation
This Program (when finished) will have 2 part
- Parsing - convert string into graph [Prase.py]
- Differentation - using graph to calculate partial derivitive [graph.py]
- - using graph to create new differentation graph [graph.py]
Find Differentation using graph
Input : dic -> dictionary for each Variable value var -> Variable the program will respect to
There's each graph type (class) for each support operation (+ - * / **) and also constance and variable
it recive the data in constructor, the operartion node usually have two parameter whose type is another graph
after the graph has fully constructed the top layer will call gradient(dic, var)
function that will
try to calculate it's gradient, most of the operation will be dependent on it's parameter graph and recursivly dive down the graph node
until it reach constance or variable