thunder1989 / melange Goto Github PK

@ckarthik1994 @Thunder1989

The current format seems not generalizable as a module in this project. You should make it more generalizable. Think that anybody who does not know your data format should be able to use it.

1

In general, this function is not generalizable. It can only digest the file format you have. You should design a function generic enough for every users to use. What you may do are

1. Design a data model for the timesereis data.
2. How to change the end user's data format into the target data model (that you designed) should be the end user's responsibility.
   1. The data model I used in the code I shared before was

   [ (uuid1, timestamp1, value1),
     (uuid2, timestamp2, value2),
   ]
   

   2. If you'd like not to include uuids in this data model, you can just remove that part and assuming the uuid is given as a parameter, but it'd be painful in bulk upload.
3. Then, in your function, you can transform the functions into the MongoDB model that you are using.
4. Change "filename" to srcid, which is the identifier of a sensor.

2

What is "s"?

3

Again, change the parameter to srcid

4

Where is this module?

5

Use consistent naming conventions. file_path. I think it could be filename as you did in the other codes. But anyway, filename should be agnostic to your implementation, so I'd say if there is any code needeing filename, it'd be wrong. Such implementation should be added outside the package (like helper code.)

6

Document your data model in the code.

@Thunder1989

While reading through your code, I found that you mix-used both spaces and tab for indentations. It's hard read in text editors. Could you fix them while you are modifying your code for the common database interface?

E.g. https://github.com/Thunder1989/melange/blob/master/inferencer/building_adapter_interface.py#L31