How to re-generate dataset df at time t2 while having df at time t1 and cross-sectional dataset df' at time t1 and t2?
I have a travel survey dataset df' collected in 2017 and also 2019. Note that individuals (households here) are not necessarily identical in 2017 and 2019 but their features are.
Dataset df' in 2017:
| household | income | size | delivery |
|---|---|---|---|
| A | 100K | 2 | 2 |
| B | 150K | 4 | 0 |
Dataset df' in 2019:
| household | income | size | delivery |
|---|---|---|---|
| C | 75K | 1 | 1 |
| D | 100K | 4 | 5 |
Now I have another travel survey df (in 2017 only) that has some features in common with df':
Dataset df in 2017:
| household | rent/own | vehicle | delivery |
|---|---|---|---|
| E | rent | Ford | 0 |
| F | own | GMC | 1 |
I am wondering how to use the relationship between samples of df' in 2017 and 2019 and also the common features between df and df' to regenerate the dataset df in 2019?
Is there any method or theoretical foundation that can assist us with doing so?
Topic data-analysis dataset machine-learning
Category Data Science