# No idea of how to handle data
**URL:** https://forum.crunchdao.com/t/no-idea-of-how-to-handle-data/470
**Category:** (ended) Causality Discovery
**Created:** [August 6, 2024, 12:54pm UTC](https://forum.crunchdao.com/t/no-idea-of-how-to-handle-data/470 "2024-08-06T12:54:51Z")
**Posts on this page:** 3
**Page:** 1
### Author:  [@permanent-nosa](https://forum.crunchdao.com/u/permanent-nosa)
#### Post date: [August 6, 2024, 12:54pm UTC](https://forum.crunchdao.com/t/no-idea-of-how-to-handle-data/470/1 "2024-08-06T12:54:51Z")
I need some advise on how to approach this competition. I am specifically talking about the handling of data. Let me upfront, I have never encountered anything to do with causal inference or discovery and no data scientist. I have been getting errors such as AttributeError: ‘dict’ object has no attribute ‘shape’ to ValueError: Data must be 1-dimensional. In fact, too numerous to mention. Is there an approach to handling this data? I look forward to someone enlightening me.
---
### Author:  [@manny-hk](https://forum.crunchdao.com/u/manny-hk)
#### Post date: [August 7, 2024, 5:14am UTC](https://forum.crunchdao.com/t/no-idea-of-how-to-handle-data/470/2 "2024-08-07T05:14:57Z")
Each of the pickle files contains a dictionary. The dictionary associates a string key (containing the name of one data example). For “X\_” files, the value associated with a key is a dataframe containing the data at each node of the graph. For “y\_” files, the value is the adjacency matrix of that graph.
---
### Author:  [@permanent-nosa](https://forum.crunchdao.com/u/permanent-nosa)
#### Post date: [August 14, 2024, 6:33am UTC](https://forum.crunchdao.com/t/no-idea-of-how-to-handle-data/470/3 "2024-08-14T06:33:45Z")
Thanks for the clarification. Your response gave me the start I needed. Thanks once again.