https://github.com/lorenzhs/fastfilter_cpp

sort by:
Revision Author Date Message Commit Date
c8577fd Prototype new Xor$ extension (Xor+$ also 2/3rds possible) Xor$ is just a terrible working title 15 January 2020, 23:54:29 UTC
b46d026 Add simulation code, for build success etc 13 January 2020, 16:28:51 UTC
1c16813 Support dumping Xor filter to binary file 13 January 2020, 16:26:00 UTC
0920acb Use little endian instead of big endian in NBitArray, for speed 07 January 2020, 17:06:10 UTC
fd1fc80 Make it easier to exclude the Linux event code I'm getting ENOENT on my dev machine for some reason, so I want to disable it. 06 January 2020, 22:39:54 UTC
3af42d3 Update README.md 26 December 2019, 16:09:47 UTC
da862c2 Update README.md 26 December 2019, 01:40:49 UTC
022414a Update README.md 23 December 2019, 17:00:08 UTC
5b6d4b5 Update README.md 23 December 2019, 15:29:39 UTC
bd43d61 Update README.md 19 December 2019, 23:47:32 UTC
7b3dc82 Benchmark for cuckoo filters at 31.5 million keys (where fpp for 2^n and not-2^n are similar) 24 August 2019, 13:01:12 UTC
af2ff2a Morton filter: load factor of 0.95, with GCC 23 August 2019, 04:59:16 UTC
3515232 Added benchmarks and a space usage test script 22 August 2019, 12:45:42 UTC
ad25fa6 Morton filter: calculate memory usage (wasn't calculated before) 22 August 2019, 11:37:11 UTC
b7a7db7 Switch back to fast select 25 June 2019, 15:41:20 UTC
8e2fa8c Test alternative (slower) select64 25 June 2019, 15:34:14 UTC
5ac09e4 Succinct counting blocked Bloom using rank instead of select: bugfix 25 June 2019, 10:27:58 UTC
c95e162 Added succinct counting blocked Bloom using rank instead of select 21 June 2019, 20:04:59 UTC
905ab96 Improve benchmark docs 02 May 2019, 13:42:53 UTC
a59ac0b Update README.md 01 May 2019, 11:05:28 UTC
0990e51 Improved benchmark instructions 01 May 2019, 10:53:42 UTC
8a4ed4c Rename benchmark shell script 01 May 2019, 10:41:31 UTC
18c8967 Ignore compiler warning 01 May 2019, 10:24:03 UTC
70a9dcb Cleanup 01 May 2019, 08:37:13 UTC
a772205 Cleanup 01 May 2019, 08:36:31 UTC
f1cfb60 Xor+ remove the option to use Poppy as an alternative to Rank9 (Poppy is slower) 01 May 2019, 08:25:11 UTC
61240bd Reorganizse filter code 01 May 2019, 08:16:58 UTC
448c70b Update README.md 29 April 2019, 21:44:57 UTC
4e905d1 More details regarding usage. 29 April 2019, 21:44:20 UTC
75e15f5 Only enabling the actually relevant algo by default. 29 April 2019, 21:32:44 UTC
a0d4d85 Minor update. 26 April 2019, 23:08:45 UTC
2caf7b5 Updating the single header version. 26 April 2019, 18:28:13 UTC
9db3b57 Add Morton filters 25 April 2019, 15:04:02 UTC
061bf98 Reorganize and add xor-singleheader 25 April 2019, 08:04:30 UTC
2b71f46 Update README.md 24 April 2019, 13:11:04 UTC
64bee75 Update README.md 19 April 2019, 21:06:56 UTC
839bb88 Add optional Poppy rank data structure (disabled by default) 18 April 2019, 20:15:14 UTC
c6a5cb5 Merge branch 'master' of github.com:FastFilter/fastfilter_cpp 10 April 2019, 21:30:03 UTC
0f8cc2c Making the private stuff public so we can serialize it as a client. 10 April 2019, 21:29:36 UTC
2b5ff71 Let us make attributes public, otherwise serializing from the outside is hard. 10 April 2019, 16:57:10 UTC
ef12a10 The seed should be public otherwise serialization is hard. 10 April 2019, 16:54:42 UTC
34e11ee Making the API less restrictive. 10 April 2019, 15:59:19 UTC
2bc6cfb Bugfix: SimdBlockFilter with 2^n size was used instead of variable size 09 April 2019, 08:48:51 UTC
dd28111 Fix AArch64 29 March 2019, 20:36:02 UTC
d026b8a New: succinct counting blocked Bloom filter 21 December 2018, 06:06:07 UTC
610b5f6 Cleanup algorithm ids 13 December 2018, 07:57:16 UTC
cd934eb Benchark removing entries 12 December 2018, 07:25:57 UTC
17f8a7b CONTAIN_ATTRIBUTES was missing for the cuckoo filter 07 December 2018, 16:42:53 UTC
d082fcd Benchmark: a bit cleanup 07 December 2018, 09:48:37 UTC
c710d24 Fix counting Bloom filter 06 December 2018, 15:35:39 UTC
ed55787 Counting Bloom filter: fix AddAll 06 December 2018, 07:49:09 UTC
29dc59b Counting Bloom filter: AddAll 06 December 2018, 07:03:45 UTC
e4aca7b Succinct counting Bloom filter: tiny improvment 05 December 2018, 16:35:13 UTC
5b5a198 Succinct counting Bloom filter: support non-BMI2 processors 05 December 2018, 07:21:02 UTC
716f774 Merge branch 'master' of https://github.com/FastFilter/fastfilter_cpp 04 December 2018, 17:06:08 UTC
603adbc New counting Bloom filter and succinct counting Bloom filter 04 December 2018, 16:45:22 UTC
2aff2fe Saving... 29 November 2018, 00:40:16 UTC
d2daaf1 Adding power9 results. 22 November 2018, 23:58:02 UTC
faa1f84 Merge pull request #1 from lemire/master Creating a viable ARM NEON implementation 22 November 2018, 13:10:35 UTC
1ab293d Pushing some reference intel perf. 22 November 2018, 02:51:07 UTC
317838b Fixing the ARM Neon implementation. 22 November 2018, 02:47:55 UTC
a25abb3 Experimenting with the ARM block bloom version. 22 November 2018, 01:55:03 UTC
41f59c5 Update README.md 08 November 2018, 20:02:53 UTC
d1e4ff1 Move files 08 November 2018, 19:58:09 UTC
819bcbe Move files 08 November 2018, 19:56:15 UTC
ee1c324 Imported old as a subtree. 08 November 2018, 19:51:33 UTC
785e850 Bloom filter: ensure branchless is only used for k >= 3 08 November 2018, 19:35:25 UTC
09d0d23 Initial commit 07 November 2018, 20:48:46 UTC
2794c29 Enabled and slightly improved Branchless Bloom 07 November 2018, 11:32:30 UTC
2042042 Merge pull request #7 from lemire/master Introducing a branchless version of bloom + tweaking the simple block… 05 November 2018, 08:29:29 UTC
da9c69e Introducing a branchless version of bloom + tweaking the simple blocked bloom so it has a more practical proba. 04 November 2018, 20:50:22 UTC
e20eec7 Merge pull request #6 from lemire/master Faster/tighter BlockedBloom64 31 October 2018, 15:45:56 UTC
0b9d948 Implemented a naive block Bloom 30 October 2018, 18:52:01 UTC
3e76710 Merge branch 'master' of github.com:lemire/xorfilter_cpp 30 October 2018, 15:25:32 UTC
db4baf9 Fix for aarch64 (ARM NEON) version. 30 October 2018, 02:32:44 UTC
ad925f6 adding 16-byte vectorized bloom version (not very good). 30 October 2018, 01:36:36 UTC
53fcec4 Merge remote-tracking branch 'upstream/master' 29 October 2018, 20:36:57 UTC
76c35e2 Faster BlockedBloom64 27 October 2018, 21:28:28 UTC
79aae0a Merge pull request #5 from lemire/master 64-byte Blocked approach has *slightly* better space usage. 27 October 2018, 18:57:46 UTC
fedb8f1 Making it more robust. 27 October 2018, 16:33:29 UTC
3b820e0 64-byte Blocked approach has *slightly* better space usage. 27 October 2018, 16:26:31 UTC
dbc3007 Merge pull request #4 from lemire/master Various minor fixes 27 October 2018, 06:31:21 UTC
a4c05a2 Presenting the Linux metrics more elegantly. Disabling one broken algo. 26 October 2018, 23:28:39 UTC
8ee775b Fixing logical error that could lead to overflows. 26 October 2018, 22:14:42 UTC
bd72df2 Adding "named" support for the new schemes + making it easier/cleaner to run smaller tests. 26 October 2018, 20:18:11 UTC
992c083 Xor 13: using 39 of 40 bits 26 October 2018, 15:52:00 UTC
1235478 Xor 10.666: a tiny bit faster 26 October 2018, 13:50:45 UTC
05cd1ec Xor 10.666: new variant, a bit slow 26 October 2018, 10:01:17 UTC
c29e59e Xor 12: fix fingerprint hash 26 October 2018, 06:14:22 UTC
abbadfa Don't use #define 26 October 2018, 05:22:51 UTC
3ded030 Faster construction for xorfilter_2 26 October 2018, 05:22:28 UTC
bc582a5 New Xor10 variant (fast, but needs more space as only 30 of 32 bits are used) 26 October 2018, 05:20:54 UTC
6200806 Statistics also for construction 24 October 2018, 20:20:07 UTC
4fabb85 Statistics also for construction 24 October 2018, 20:14:59 UTC
7449f3e Blocked Bloom filter: bulk insert (faster only in the 100m case, and only about 10%) 24 October 2018, 20:01:33 UTC
9bc380d Blocked Bloom filter: bulk insert (faster only in the 100m case, and only about 10%) 24 October 2018, 19:58:57 UTC
e7ef113 Blocked Bloom filter: bulk insert (faster only in the 100m case, and only about 10%) 24 October 2018, 19:53:50 UTC
d902530 Merge pull request #3 from lemire/master Greatly accelerate the benchmarks + allow to provide a list of ids (comma-separated) 24 October 2018, 19:53:38 UTC
56b7dc8 Moving up the construction of the samples so that we can greatly accelerate the benchmarks when you are testing different algos. 24 October 2018, 17:22:05 UTC
e3d391c Slow random number generation is totally unwarranted. 24 October 2018, 15:23:22 UTC
back to top