Generic implementation of optimal grouping of objects using dynamic programming #272
Conversation
|
The algorithm will work on an array of objects if I am not wrong. So, the code should be shifted to |
|
I have moved the file to linear_data_structures. |
|
Hi, |
|
Hi, |
|
Add some tests for your patch. See |
|
|
||
|
|
||
|
|
||
|
|
||
|
|
| """ | ||
| gets a value | ||
| """ | ||
| return matrix[lookup_index[0]][lookup_index[1]] |
There was a problem hiding this comment.
One liner functions aren't recommended. They just slow down the code.
There was a problem hiding this comment.
See my suggestions at https://github.com/codezonediitj/pydatastructs/pull/272/files#r426157438
You have to do something similar with every one liner function.
| """ | ||
|
|
||
| # gets the present value at the present index | ||
| present_value = _get_value_opt_group(cost_storage, lookup_index) |
There was a problem hiding this comment.
| present_value = _get_value_opt_group(cost_storage, lookup_index) | |
| present_value = cost_storage[lookup_index[0]][lookup_index[1]] |
| def _test_lookup_function(lookup_index: List[int], input_index: List[int]): | ||
| if lookup_index is None: | ||
| raise TypeError( | ||
| 'Check lookup_function: returning wrong type should return an array of start and end index') | ||
|
|
||
| if lookup_index.__len__() < 2: | ||
| raise ValueError( | ||
| 'Check lookup_function:lookup index should at least have 2 integer items, first specifying the start and second specifying the last indices') |
There was a problem hiding this comment.
| if input_index == lookup_index: | ||
| raise RuntimeError( | ||
| 'Check lookup_function:verify get_lookup_fn giving same output as input which will lead to infinite loop') |
There was a problem hiding this comment.
| if input_index == lookup_index: | |
| raise RuntimeError( | |
| 'Check lookup_function:verify get_lookup_fn giving same output as input which will lead to infinite loop') | |
| assert input_index == lookup_index |
| if get_lookup_fn is None or type(get_lookup_fn) is not FunctionType: | ||
| raise TypeError( | ||
| 'get_lookup_fn cannot be none and should be a function with 3 arguments') |
There was a problem hiding this comment.
Do not worry about type checking. If user will give garbage, they will get garbage.
| @@ -3,13 +3,16 @@ | |||
| from pydatastructs.utils.misc_util import _check_type, _comp | |||
| from concurrent.futures import ThreadPoolExecutor | |||
| from math import log, floor | |||
| from types import * | |||
There was a problem hiding this comment.
Do not do *, import only those things which you want to use. Avoid namespace pollution.
Codecov Report
@@ Coverage Diff @@
## master #272 +/- ##
=============================================
- Coverage 98.842% 98.662% -0.181%
=============================================
Files 23 23
Lines 2420 2467 +47
=============================================
+ Hits 2392 2434 +42
- Misses 28 33 +5
|
|
Hi, |
|
@pravalikavis Please resolve merge conflicts. |
|
My only concern with dynamic programming problems is that they aren't that abstract and generalized. Moreover, they are pseudo-polynomial solutions to optimisation problems (mostly discrete). So possibly, we can add a separate optimisation module and put the dynamic programming solutions there. In fact they can also contain parametrised approximation algorithm. Anyways, we should first check if there is already a python package for solving these kind of formulations. |
|
Agreed on this |
References to other Issues or PRs or Relevant literature
fixes #230
Brief description of what is fixed or changed
Other comments