CODEWITHSUNDEEP
data-sciencecluster-analysis
Aiden Blake
Aiden Blake

Reputation: 117

Suitable Clustering Approach

I've got a total of 9 sensors in the ground, which measure the water content of the soil. 1-3 are in a depth of 1m, 4-6 are in a depth of 2m and sensors 7-9 are in a depth of 3m.

My dataset also contains the precipiation of the location. It is hourly data:

Time Sensor-ID Precipitation Soil Water Content
2022-01-01 11:00 1 74 120
2022-01-01 11:00 2 74 100
2022-01-01 11:00 3 74 110
... ... ... ...
2022-01-01 11:00 9 74 30

The goal now is to find out if the different ground / soil depths behave differently regarding the water content after raining (over time).

I thought about a clustering method to find out if the sensors can be clustered based on the data and confirm this. Since I'm not very experienced in data science, would that be the right approach and is it even possible to analyse it with clustering?

Upvotes: 0

Views: 19

Answers (1)

Farjad Ahmad Mir
Farjad Ahmad Mir

Reputation: 11

For clustering, you can add a new column with three new classes to your data - for 1-3 sensors : Class 1, for 4-6 sensors : Class 2, for 7-9 sensors : Class 3 and perform your analysis using the new classes. Either can be done using Python, Power BI or Excel.

You should start by analyzing different variables w.r.t to the sensors at different ground depths: Use univariate, Bi-Variate and Multi-Variate plots to derive your goal.

Upvotes: 1

Related Questions