[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15581416#comment-15581416
]
Yun Ni commented on SPARK-5992:
---
Yes, I have implemented cosine, jaccard, euclidean and hamming distance.
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15581412#comment-15581412
]
Yun Ni commented on SPARK-5992:
---
Yes, I do have comparison with full scan. Usually kNN with 1 hash function
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15581366#comment-15581366
]
Debasish Das commented on SPARK-5992:
-
Also do you have hash function for euclidean distance? We use
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15581361#comment-15581361
]
Debasish Das commented on SPARK-5992:
-
Did you compare with brute force knn ? Normally lsh does not
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15581345#comment-15581345
]
Yun Ni commented on SPARK-5992:
---
Some tests on the WEX Open Dataset, with steps and results:
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15503971#comment-15503971
]
Apache Spark commented on SPARK-5992:
-
User 'Yunni' has created a pull request for this issue:
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15502493#comment-15502493
]
Yun Ni commented on SPARK-5992:
---
Hi Joseph,
I have made an initial PR based on the design doc:
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15478135#comment-15478135
]
Yun Ni commented on SPARK-5992:
---
Thank you very much for reviewing it, Joseph!
I will work on the first
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15478125#comment-15478125
]
Joseph K. Bradley commented on SPARK-5992:
--
The design doc LGTM! Thanks for updating it. Shall
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15453980#comment-15453980
]
Yun Ni commented on SPARK-5992:
---
Thanks very much for reviewing, Joseph!
Based on your comments, I have
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15451091#comment-15451091
]
Joseph K. Bradley commented on SPARK-5992:
--
Awesome, thanks! I made some comments, but it looks
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15432186#comment-15432186
]
Yun Ni commented on SPARK-5992:
---
Hi,
We are engineers from Uber. Here is our design doc for LSH:
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15400751#comment-15400751
]
snehil suresh wakchaure commented on SPARK-5992:
Hello, just curious to know if I can
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15389321#comment-15389321
]
tony commented on SPARK-5992:
-
Hi,
I am new to spark. How can I make a contribution to Spark LSH
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15345890#comment-15345890
]
Xiangrui Meng commented on SPARK-5992:
--
FYI, I had an offline discussion with Kelvin Chu and Erran Li
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15159679#comment-15159679
]
Karl Higley commented on SPARK-5992:
I've been working on [a Spark package for approximate nearest
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15029318#comment-15029318
]
Karl Higley commented on SPARK-5992:
I'm a bit confused by this section of the design doc:
{quote}
It
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15010635#comment-15010635
]
Jonas Seiler commented on SPARK-5992:
-
Very nice!
Hope there will be soon a common interface and we
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14904500#comment-14904500
]
Özgür Demir commented on SPARK-5992:
hi, we just open sourced an lsh topk implementation for spark.
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14721066#comment-14721066
]
Maruf Aytekin commented on SPARK-5992:
--
I have developed spark implementation of LSH
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14637337#comment-14637337
]
Maruf Aytekin commented on SPARK-5992:
--
In addition to Charikar's scheme for cosine
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14574933#comment-14574933
]
Yu Ishikawa commented on SPARK-5992:
h2. Initial version of desing doc
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14575002#comment-14575002
]
Karl Higley commented on SPARK-5992:
To make it easier to define a common interface,
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14574750#comment-14574750
]
Yu Ishikawa commented on SPARK-5992:
[~debasish83] Thank you for your comment. I
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14575079#comment-14575079
]
Joseph K. Bradley commented on SPARK-5992:
--
I'm just noting here that
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14512843#comment-14512843
]
Debasish Das commented on SPARK-5992:
-
Did someone compared algebird LSH with spark
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14395415#comment-14395415
]
Yu Ishikawa commented on SPARK-5992:
Yes, I will try to write a design doc. Thank you
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14392366#comment-14392366
]
Yu Ishikawa commented on SPARK-5992:
As I said when we talked in Spark Summit East, I
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14357335#comment-14357335
]
Joseph K. Bradley commented on SPARK-5992:
--
I'm not sure what the most important
[
https://issues.apache.org/jira/browse/SPARK-5992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14356507#comment-14356507
]
Yu Ishikawa commented on SPARK-5992:
What kind of LSH algorithms should we support at
30 matches
Mail list logo
