Hi, i have a problem with my time series data in Timelion.
For example, I have a set of cars, that have index 'car' and every document besides different metadata has the most important for me fields: price (i may collect several price value for the data) , model (e.g. Mercedes) and date (the appropriate data of the price for the model was got from the market)
I want to gain time series data to see the price of the specific (e.g. only for Mercedes) car for this date,
I started to use this query:
.es(index=car, q='model:Mercedes', timefield='date', metric='max:price').color(ff0000)
then i try to change the model and watch the difference:
.es(index=car, q='model:BMW', timefield='date', metric='max:price').color(ff0000)
and i get the same chart, that is not true.
after this i tried to split the models in one chart:
.es(index=car,split='model:5', timefield='date', metric='max:price').color(#ff0000)
and i get only one line.
The thing that a i want to do may be got from the if conditional, but the if operator must be number or series, but i need the String: .es(q='model').if(eq,Mercedes,.es(timefield='date', metric='max:price')) but it couldn't be done.
Of course, i 've checked that the data for different models is defferent and exist.
Please, help me to get this data like i've described to you.
I assume that data is supposed to mean date in this example? In order to split on the field model it would have to be indexed as a keyword. If you enter the queries model:Mercedes and model:BMW into the Discover search bar, do you see the correct subset of documents?
I tried reproducing the problem with the information I got from you so far, but failed. If you split on model, what is the label of the one series that is displayed?
I will have 5 different labels, but all of them return the one line, like they are the same prices for all of the models, that is not right. i may hide every label and don't notice the changes, i will notice them only if i hide all of them.
Sorry, I misunderstood you there. So you get 5 identical series, which means that for each bucket the maximum price is the same.
To give me a better idea of how the values are distributed, would it be possible to run the following query after substituting the <FROM> and <TO> placeholders for the timestamps of the interval you are looking at in timelion and post the results?
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.