Context Navigation

Changes between Version 62 and Version 63 of DatabaseBasedAnalysis

-              v62
+              v63
 I have kept that here for convenience (and with the current server, it works, simply because the optimizer does not shuffle the columns.
 Now let's have a look at how the query should look correctly ('''Don't panic!'''):
+Now let's have a look at how the query should look correctly (''' Don't panic! '''):
 {{{#!sql
 …
 }}}
 Why is this so much faster? I guess it has two reasons:
+Note that this is the result of the analysis of a dataset which, as a whole, is more than 100GB and even reduced to Crab after quality cuts still is of the order of GB! Why is this so much faster? I guess it has two reasons:
 ) Only variables which are really needed in the next step are kept (reduces memory consumption)
 …
 Another example (apart from the obvious difficulty to read the query) is that it is very well structured. For each step it is very well defined what values
 are available from the previous step (the fields which are requested in each SELECT statement) and it is very well defined in which order the algorithm is executed.
+If somebody find a way to write that query in an easier readable way, I would be glad to know.