so we should perform this one step on line53, and then generate highlevel data for this model (#88)
2015-08-11 22348, 2015
alastairp
and we will then have two pieces of information: the ground truth in the dataset editor; and the high-level estimates from running our model
2015-08-11 22314, 2015
alastairp
I have code which will give an overall accuracy of the model based on these two bits of information
2015-08-11 22332, 2015
alastairp
if you got it to this stage, it'd be a full dataset/evaluation cycle
2015-08-11 22302, 2015
Gentlecat
alastairp: do we need to generate history file during every evaluation?
2015-08-11 22323, 2015
Gentlecat
and where should we store it?
2015-08-11 22344, 2015
ariscop joined the channel
2015-08-11 22332, 2015
alastairp
yeah, I'm not sure at what stage we should generate it
2015-08-11 22323, 2015
alastairp
if we wait, then the evaluation / parameter selection will happen faster
2015-08-11 22326, 2015
alastairp
since there is one less step
2015-08-11 22300, 2015
alastairp
so we could not do it during every evaluation; only when someone is happy with their accuracy
2015-08-11 22321, 2015
alastairp
you could just store it in a data directory
2015-08-11 22333, 2015
alastairp
data/models/uuid/date/file.history
2015-08-11 22354, 2015
Gentlecat
ok
2015-08-11 22356, 2015
Gentlecat
uuid of evaluation job?
2015-08-11 22307, 2015
alastairp
ah, is there uuid for dataset and for evaluation?
2015-08-11 22316, 2015
alastairp
(I was thinking of dataset)
2015-08-11 22335, 2015
alastairp
but if there is one for evaluation, use that. you don't need a date then (since you can get that data from the database)
2015-08-11 22300, 2015
johtso joined the channel
2015-08-11 22354, 2015
D4RK-PH0ENiX joined the channel
2015-08-11 22348, 2015
CatQuest
uhm. am I doing something wrong? release editor doesn't go through. any change and clicking "enter edit" and it spins for a while and then goes akc ot normal but no submitting is done? tired both beta and regular
2015-08-11 22344, 2015
Gentlecat
alastairp: do you know how long this step is?
2015-08-11 22328, 2015
Gentlecat
I think having that running separately makes things unnecessarily more complicated
2015-08-11 22306, 2015
Gentlecat
we probably need another queue for generating history files
2015-08-11 22302, 2015
Gentlecat
evaluation script will need significant changes if we are going to reuse it for this task
2015-08-11 22330, 2015
alastairp
It's basically the same length as a parameter permutation
I don't think I even changed anything. it's the same VM, same dataset
2015-08-11 22342, 2015
alastairp
hmm
2015-08-11 22345, 2015
alastairp
did you update gaia?
2015-08-11 22355, 2015
alastairp
because I made a change recently to the python bindings
2015-08-11 22325, 2015
Gentlecat
I don't think so
2015-08-11 22345, 2015
Gentlecat
I'll try to set it up again
2015-08-11 22336, 2015
alastairp
I'll try a few demos - we should just be able to load up the python interface in a shell to see where the problem is
2015-08-11 22347, 2015
alastairp
I'll revert my change and see if that was the problem
2015-08-11 22303, 2015
alastairp
ahhhh
2015-08-11 22316, 2015
alastairp
Gentlecat: double-check that your class names are strings, not uuid objects
2015-08-11 22345, 2015
alastairp
in fact, strings. not anything else
2015-08-11 22320, 2015
alastairp
you should be able to reproduce it in a python shell with p = gaia2.Point(); p[9] = "foo"
2015-08-11 22337, 2015
alastairp
whereas p["string"] = "foo" will give a more normal error
2015-08-11 22332, 2015
Gentlecat
let me check
2015-08-11 22300, 2015
Gentlecat
I think it's a unicode string
2015-08-11 22323, 2015
Gentlecat
this raises the same exception: p[unicode("hello")] = "test"
2015-08-11 22331, 2015
alastairp
ah!
2015-08-11 22333, 2015
alastairp
there we go
2015-08-11 22345, 2015
alastairp
1) open ticket in gaia
2015-08-11 22350, 2015
alastairp
2) convert to str
2015-08-11 22346, 2015
Gentlecat
alastairp: wait, it's different function
2015-08-11 22315, 2015
Gentlecat
these examples reference Point_setValue function
2015-08-11 22337, 2015
alastairp
I just changed your create_groundtruth method
2015-08-11 22340, 2015
alastairp
line 91
2015-08-11 22344, 2015
alastairp
str(cls["name"])
2015-08-11 22345, 2015
Gentlecat
I mean Point_setLabel
2015-08-11 22345, 2015
alastairp
fixed
2015-08-11 22304, 2015
Gentlecat
maybe it's better to use .encode("UTF-8")?
2015-08-11 22316, 2015
alastairp
yeah, that should work just as well
2015-08-11 22329, 2015
alastairp
do we have any restrictions on label names?
2015-08-11 22343, 2015
alastairp
perhaps we should actually limit them to [a-zA-Z0-9]+
2015-08-11 22357, 2015
Gentlecat
no, not really
2015-08-11 22342, 2015
Gentlecat
yeah, that's one way to do it, but seems hacky
2015-08-11 22310, 2015
alastairp
I don't mind if you str() or .encode
2015-08-11 22326, 2015
alastairp
oh - I bet it worked for you before you set the default encoding of the database connection, right? :)
2015-08-11 22347, 2015
Gentlecat
right!
2015-08-11 22301, 2015
Gentlecat
that's what it was :)
2015-08-11 22309, 2015
alastairp
OK, cool. I was trying to fix this issue before I gave you some more feedback on the website interface
2015-08-11 22321, 2015
alastairp
now that I've succesfully done a model, I'll do this for you too
2015-08-11 22330, 2015
Gentlecat
feedback \o/
2015-08-11 22321, 2015
ruchiranga joined the channel
2015-08-11 22359, 2015
ruchiranga
how should a query to the web service be written with something like 7" vinyl ? for example /ws/2/release?fmt=json&query=format:(7" vinyl)&offset=0 does not work because of the quotation mark
2015-08-11 22355, 2015
alastairp
can you urlencode it? " is %22
2015-08-11 22322, 2015
alastairp
Gentlecat: did you open a ticket in gaia for non-string values?
2015-08-11 22344, 2015
Gentlecat
not yet, it's on the list
2015-08-11 22355, 2015
alastairp
ok, cool
2015-08-11 22304, 2015
ruchiranga
alastairp: tried it but the same error comes
2015-08-11 22307, 2015
ruchiranga
{"error":"Search server could not complete query: Bad request"}
