hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Dalia Sobhy <dalia.mohso...@hotmail.com>
Subject Re: Question
Date Wed, 14 Nov 2012 09:09:26 GMT
String

Sent from my iPhone

On 2012-11-14, at 12:30 AM, "Ted Yu" <yuzhihong@gmail.com> wrote:

> What number format did you use for the cardiac column family ?
> I assume you store double as byte array.
> 
> Cheers
> 
> On Tue, Nov 13, 2012 at 2:14 PM, Dalia Sobhy <dalia.mohsobhy@hotmail.com>wrote:
> 
>> Hello Ted:
>> 
>> I have a patient table which contains two column families: info and
>> cardiac.
>> 
>> I need to count the number of people who suffer from a particular
>> diagnosis within cardiac.
>> 
>> For instance the number of patients suffering from "heart failure"
>> 
>> Sent from my iPad
>> 
>> On Nov 13, 2012, at 11:00 PM, "Ted Yu" <yuzhihong@gmail.com> wrote:
>> 
>>> Currently http://hbase.apache.org/book.html gives me networks error.
>>> You should be able to find example of how Java client connects to HBase
>> cluster.
>>> 
>>> How ColumnInterpreter is initialized depends on your schema. Can you
>> tell us more about your schema ?
>>> 
>>> Thanks
>>> 
>>> On Tue, Nov 13, 2012 at 12:40 PM, Dalia Sobhy <
>> dalia.mohsobhy@hotmail.com> wrote:
>>>> 
>>>> 
>>>> 
>>>> 
>>>> Begin forwarded message:
>>>> 
>>>>> From: Dalia Sobhy <dalia.mohsobhy@hotmail.com>
>>>>> Date: November 12, 2012, 10:12:57 PM GMT+02:00
>>>>> To: <yuzhihong@gmail.com>
>>>>> Subject: RE: Question
>>>>> 
>>>>> Thanks,
>>>>> 
>>>>> But how to initialize the ColumnInterpreter, I mean how to link it
>> with a table.
>>>>> 
>>>>> Another aspect, its my first time to use java classes for hbase, I was
>> using thrift server to connect to hbase. But due to the need for aggregate
>> functions i shifted to hbase 0.92.
>>>>> 
>>>>> So could you explain to me in brief how are the java classes connected
>> to hbase server.
>>>>> 
>>>>> Thanks in advance :D
>>>>> 
>>>>> Date: Mon, 12 Nov 2012 11:35:58 -0800
>>>>> 
>>>>> Subject: Re: Question
>>>>> From: yuzhihong@gmail.com
>>>>> To: dalia.mohsobhy@hotmail.com
>>>>> CC: user@hbase.apache.org
>>>>> 
>>>>> 
>>>>> double is primitive type.
>>>>> Replace with Double.
>>>>> 
>>>>> Cheers
>>>>> 
>>>>> On Mon, Nov 12, 2012 at 11:31 AM, Dalia Sobhy <
>> dalia.mohsobhy@hotmail.com> wrote:
>>>>> 
>>>>> 
>>>>> Syntax error on token "double", Dimensions expected after this token
>>>>> 
>>>>> Pastebin link:
>>>>> http://pastebin.com/dcvaMvLw
>>>>> 
>>>>> I am trying to send it on the group but I dunno why hotmail disagrees
>> and sends me postmaster failure so weird!!!
>>>>> 
>>>>> 
>>>>>> CC: user@hbase.apache.org
>>>>>> From: yuzhihong@gmail.com
>>>>>> Subject: Re: Question
>>>>>> Date: Sun, 11 Nov 2012 12:31:16 -0800
>>>>>> To: user@hbase.apache.org
>>>>> 
>>>>>> 
>>>>>> What error did you get ?
>>>>>> Please use pastebin to share.
>>>>>> 
>>>>>> Thanks
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> On Nov 11, 2012, at 11:42 AM, Dalia Sobhy <
>> dalia.mohsobhy@hotmail.com> wrote:
>>>>>> 
>>>>>>> 
>>>>>>> Hi Ted,
>>>>>>> 
>>>>>>> I am using hbase 0.92.1
>>>>>>> 
>>>>>>> When I tried the following code I got errors:
>>>>>>> 
>>>>>>> AggregateProtocol agg;
>>>>>>> long rowcount = agg.getRowNum(ColumnInterpreter<string, int>,
>> scan);
>>>>>>> 
>>>>>>> double sum = aggregationClient.avg(TABLE_NAME, ci, scan);
>>>>>>> 
>>>>>>> even though I tried it with String (s in capital) same errors,
do
>> u know why?
>>>>>>> 
>>>>>>>> Date: Sun, 11 Nov 2012 08:24:29 -0800
>>>>>>>> Subject: Re: Question
>>>>>>>> From: yuzhihong@gmail.com
>>>>>>>> To: user@hbase.apache.org
>>>>>>>> 
>>>>>>>> Dalia:
>>>>>>>> If you look at:
>>>>>>>> public interface ColumnInterpreter<T, S> {
>>>>>>>> 
>>>>>>>> you would see:
>>>>>>>> * @param <T> Cell value data type
>>>>>>>> * @param <S> Promoted data type
>>>>>>>> 
>>>>>>>> S can represent sum of values of T. The rationale was that
>> aggregation of
>>>>>>>> one type (integer, e.g.) may have to be represented by another
>> type
>>>>>>>> (double, e.g.)
>>>>>>>> 
>>>>>>>> From this blog you can get some history on this subject:
>> http://zhihongyu.blogspot.com/2011/03/genericizing-endpointcoprocessor.html
>>>>>>>> 
>>>>>>>> For row counting, it is already implemented:
>>>>>>>> public <T, S> long getRowNum(ColumnInterpreter<T,
S> ci, Scan
>> scan)
>>>>>>>> You can embed your filter in scan object.
>>>>>>>> 
>>>>>>>> Cheers
>>>>>>>> 
>>>>>>>> On Sun, Nov 11, 2012 at 8:07 AM, Dalia Sobhy <
>> dalia.mohsobhy@hotmail.com>wrote:
>>>>>>>> 
>>>>>>>>> 
>>>>>>>>> 
>>>>>>>>> Thanks Ted :D
>>>>>>>>> 
>>>>>>>>> But can anyone explain to me this sentence:
>>>>>>>>> 
>>>>>>>>> ColumnInterpreter<T, S> ci
>>>>>>>>> 
>>>>>>>>> How to set a ColumnInterpreter ?? What is T and Whats
S?
>>>>>>>>> 
>>>>>>>>> I want to perform average on an Aggregationclient, so
could
>> anyone provide
>>>>>>>>> me with a sample code?
>>>>>>>>> 
>>>>>>>>> Also if i want to count the number of rows which has
a specific
>> value ?
>>>>>>>>> What to use Filters or AggregationClient?
>>>>>>>>> 
>>>>>>>>> Thanks in advance,
>>>>>>>>> 
>>>>>>>>>> Date: Sat, 10 Nov 2012 14:25:22 -0800
>>>>>>>>>> Subject: Re: Question
>>>>>>>>>> From: yuzhihong@gmail.com
>>>>>>>>>> To: user@hbase.apache.org
>>>>>>>>>> 
>>>>>>>>>> Take a look at AggregationClient.java and
>> TestAggregateProtocol.java
>>>>>>>>>> 
>>>>>>>>>> Cheers
>>>>>>>>>> 
>>>>>>>>>> On Sat, Nov 10, 2012 at 2:11 PM, Dalia Sobhy <
>> dalia.mohsobhy@hotmail.com
>>>>>>>>>> wrote:
>>>>>>>>>> 
>>>>>>>>>>> 
>>>>>>>>>>> Dear all,
>>>>>>>>>>> 
>>>>>>>>>>> I want to use the aggregate functions, but do
anyone have a
>> code or
>>>>>>>>> sth to
>>>>>>>>>>> understand them ..
>>>>>>>>>>> 
>>>>>>>>>>> Thanks
>>>>>>>>>>> 
>>>>>>>>>>> Best Regards
>>>>>>>>>>> Dalia
>> 

Mime
View raw message