Find similar record in the dataset

[janika]

I have two data sets:

historical weather and day specific data:

Year, month, date, hour (in three hour brackets), clouds, humidity, pressure, temperature, weather_id, wind_deg, wind speed, weekday 

2018 9 21 18 75 100 1011 19 0 500 320 6.7 5
2018 9 22 12 75 69 1009 22 0 520 50 2.1 6
2018 9 22 15 75 83 1007 23 0 520 80 2.1  6

....

2019 11 27 18 75 100 1011 19 0 500 320 6.7 5
2019 11 27 12 75 69 1009 22 0 500 50 2.1 6
2019 11 27 15 75 83 1007 23 0 490 80 2.1  6

For the 5 day weather forecast I want to find a most similar weather on the same weekday in the historical data.

2 outputs:

1) Year, month, date for the closest similar day in the historical weather

2) Year, month, date, hour for the closest similar day in the historical weather only. 

Any ideas are most welcome 🙂

[janika · Accepted solution]

I think I found a solution by using Find Nearest Neighbor.