gini_impurity function property and test cases for this #308
Conversation
az85252
requested review from
AnhTruong,
ChrisWallace2020,
grant-eden,
JGSweets and
lettergram
as code owners
| profile["statistics"].update( | ||
| dict(gini_impurity=self.gini_impurity) | ||
| ) |
There was a problem hiding this comment.
I believe you could just add the gini_impurity key to the call to update above, with the categories.
| df_categorical = pd.Series(["y", "y", "y", "y", "n", "n", "n"]) | ||
| profile = CategoricalColumn(df_categorical.name) | ||
| profile.update(df_categorical) | ||
| expected_val = ((4 / 7) * (3/7)) + ((4 / 7) * (3/7)) |
There was a problem hiding this comment.
Make the spacing spacing on (4 / 7) consistent with (3/7)
| profile["statistics"].update( | ||
| dict(categories=self.categories) | ||
| ) |
There was a problem hiding this comment.
not your issue, but if you have to make another change in this pr, could you change this to match the format below:
profile["statistics"]["categories"] = self.categories
| def gini_impurity(self): | ||
| """ | ||
| Property for gini impurity |
There was a problem hiding this comment.
lets put a reference to the formula or show the formula used here. Also need a space between the description and the return.
| return None | ||
| summation = 0 | ||
| total = sum(self._categories.values()) |
There was a problem hiding this comment.
this value may already exist as self.sample_size
| expected_gini = .914600550 | ||
| self.assertAlmostEqual(report_gini, expected_gini) |
There was a problem hiding this comment.
is it possible to get the exact math like above or do we have to do almost equal? if we can get it then we won't need to pop it either.
| print(self.sample_size) |
There was a problem hiding this comment.
remove the print and comment
| if self.sample_size == 0: | ||
| return None | ||
| summation = 0 |
There was a problem hiding this comment.
Maybe just call this gini_impurity? thoughts?
* added gini_impurity function property and test cases for this * fixed documentation for gini impurity * fixed syntax and test cases related to gini_impurity * edited test cases and code related to gini_impurity * deleted extra code and simplified variable names
I have added gini impurity into categorical column, basically likelihood of an incorrect classification of a new instance of a random variable. Please review.