set row value from data

[goldalworming]

Hi, I'm a bit confuse about display data in pivottable
rows: [
{
NO: "1",
NAME: "2BKS000238/2BKS-JTN0101-0810BI",
VENDOR: "Research In Motion",
CNT_CLIENT: "11",
HIT_AUTH: "25",
HIT_ASOC: "0",
CLIENT_AUTH: "11",
CLIENT_ASOC: "0"
},
{
NO: "2",
NAME: "2BKS000238/2BKS-JTN0101-0810BI",
VENDOR: "Liteon",
CNT_CLIENT: "1",
HIT_AUTH: "1",
HIT_ASOC: "0",
CLIENT_AUTH: "1",
CLIENT_ASOC: "0"
},
{
NO: "3",
NAME: "2BKS000238/2BKS-JTN0101-0810BI",
VENDOR: "Unknown",
CNT_CLIENT: "5",
HIT_AUTH: "5",
HIT_ASOC: "0",
CLIENT_AUTH: "5",
CLIENT_ASOC: "0"
}
]

is it possible to create table like this

----------------------------CNT_CLIENT | HIT_AUTH | HIT_ASOC | CLIENT_AUTH | CLIENT_ASOC
Research In motion 11 | 25 | 0 | 11 | 0
Liteon 1 | 1| 0 | 1 | 0
Unknown 5 | 5 | 0 | 5 | 0

[nicolaskruchten]

It will not be possible to build such a table with this library unfortunately. The types of tables you would be able to build would have, say, the values of the Vendor attribute down the side like you do here, but then the values of, say, the CNT_CLIENT attribute across the top, and a single aggregated value in each cell.

What you are looking for would involve having multiple aggragators, which is an often-request feature.

[goldalworming]

how many number of record pivottable could handle so count work fine without performance issue??

how about implement #30 ?

[nicolaskruchten]

I doubt that #30 will be implemented particularly soon, to be honest.

Regarding the performance question, it depends on the nature of the data, how you're loading it and your hardware, but I would say up to a few tens of thousands of rows is safe, and a hundred thousand is close to the limit. Your mileage may very.

[goldalworming]

ok, but display so much data would bad in load data..
I hope multiple aggregators would be implemented as soon as possible..

thanks