This time, we are looking to get more details than the first two questions. We want to know how the temperature and humidity vary at each location over time. We will solve it for temperature. 

Go to the Visualizations tab in Kibana and create a new Line chart visualization, the same as before:

1. Add a filter for sensorType: Temperature as we did before.
2. Set up the Metrics section exactly same as the first chart that we created, that is Average Temperature over time on the reading field.
3. Since we are aggregating the data over the time field, we need to choose the Date Histogram aggregation in the Buckets section. Here, we should choose the time field and leave the aggregation Interval as Auto. 
4. Up to this point, this visualization is the same as Average temperature over time. We don't just want to see the average temperature over time; we want to see it per locationOnFloor, which is our most fine-grained unit of identifying a location. This is why we are splitting the series using the Terms aggregation on the  locationOnFloor in this step. We select Order By as metric: Average Temperature, keep Order as Descend, and Size to be 5 to retain only the top five locations.

We have now built a visualization that shows how the temperature changes for each value of locationOnFloor field in our data. You can clearly see that there is a spike in O-201 on December 1, 2017 at 15:00 IST. Because of this spike, we had seen the average temperature in our first visualization spike at that time. This is an important insight that we have uncovered. Save this visualization as Temperature at locations over time.

A visualization for humidity can be created by following the same steps but just replacing Temperature with Humidity.