Commit Graph

28 Commits

Author SHA1 Message Date
Amos Bird
8736a3908f left pad 2018-12-24 23:51:06 +08:00
Alexey Milovidov
ca52d96e09 Insignificant changes for PVS-Studio [#CLICKHOUSE-3] 2018-06-04 22:22:27 +03:00
Alexey Milovidov
b198766dad Simplification of aggregate functions: development [#CLICKHOUSE-2]. 2017-12-21 05:08:25 +03:00
Alexey Milovidov
ec3ec48309 Simplification of aggregate functions: development [#CLICKHOUSE-2]. 2017-12-21 05:08:25 +03:00
Alexey Milovidov
6fe1b175a0 Simplification of aggregate functions: development [#CLICKHOUSE-2]. 2017-12-21 05:08:25 +03:00
Alexey Milovidov
c990740b33 Simplification of aggregate functions: development [#CLICKHOUSE-2]. 2017-12-21 05:08:25 +03:00
Alexey Milovidov
e25514d28a Addition to the fix of generic variant of topK aggregate function [#CLICKHOUSE-2]. 2017-12-19 20:21:47 +03:00
Alexey Milovidov
b83142ea62 Fixed error found by M.Filimonov (generic implementation of topK aggregate function was totally wrong) [#CLICKHOUSE-2]. 2017-12-19 00:12:15 +03:00
Alexey Milovidov
f95ddfce10 Make DataTypes totally immutable [#CLICKHOUSE-2]. 2017-12-18 04:11:48 +03:00
Alexey Milovidov
7284a64740 Better semantic of sharing columns: development [#CLICKHOUSE-2]. 2017-12-16 00:32:25 +03:00
Alexey Milovidov
52088c9c07 Various fixes [#CLICKHOUSE-2] 2017-12-09 19:17:37 +03:00
Alexey Milovidov
c748493a86 Better [#CLICKHOUSE-2]. 2017-12-02 05:47:12 +03:00
Alexey Milovidov
7196602e12 Better [#CLICKHOUSE-2]. 2017-12-02 00:51:50 +03:00
proller
0d3c4f4d67 Core/FieldVisitors.h -> Common/FieldVisitors.h
Core/Progress.h -> IO/Progress.h
    tests: sudo --non-interactive
    Fix freebsd link
2017-11-24 16:55:31 +03:00
Marek Vavruša
927c0f7b74 dbms: Partially revert previous change
Instead of trying to convert long/unsigned long to Poco types, just
don't use them.
2017-10-13 23:46:10 -07:00
Marek Vavruša
8a0571af24 dbms: Better fix for missing unsigned long to Poco::UInt64 alias 2017-10-13 21:39:14 -07:00
Marek Vavruša
70543200be dbms: Fix build with clang after Poco update
In the current version of Poco, unsigned long no longer aliases to
UInt64 with LP64. The size_t aliases to unsigned long with clang,
so all the uses of size_t instead of UInt64 when interacting with
Poco interfaces are gone. I replaced uses with UInt64 where it makes
sense, and added an overloaded function for readVarUInt() to support size_t.
2017-10-13 18:26:24 -07:00
Alexey Milovidov
1405415f94 Miscellaneous [#CLICKHOUSE-3]. 2017-10-09 01:53:38 +03:00
alexey-milovidov
2dfb3fa463 Update AggregateFunctionTopK.h 2017-10-09 01:12:38 +03:00
alexey-milovidov
6e5ea5cc10 Update AggregateFunctionTopK.h 2017-10-09 01:12:38 +03:00
Marek Vavruša
28bb5e25cf AggregateFunctionTopK: read alphaMap for generic
* the alpha_map vector always (de)serialises
  the actual version (could empty sometimes)
* AggregateFunctionTopK generic variant deserialises
  it as well instead of ignoring it
* AggregateFunctionTopK generic variant clears the
  array before deserialising

refs #1283
2017-10-09 01:12:38 +03:00
Alexey Milovidov
ff1598c8d1 Improved performance of dynamic compilation [#CLICKHOUSE-2]. 2017-09-17 23:22:39 +03:00
alexey-milovidov
d4d266fa18 ColumnConst unification (#1011)
* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* Fixed error in ColumnArray::replicateGeneric [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].

* ColumnConst: unification (incomplete) [#CLICKHOUSE-3150].
2017-07-21 09:35:58 +03:00
Marek Vavruša
e189c39056 SpaceSaving: internal storage for StringRef{}
The SpaceSaving has now specialised storage for
some keys, which only copies keys that
are to be retained in the structure, not all.

Most of the PODs implement this interface empty,
so there shouldn’t be any extra cost.
2017-06-26 21:16:13 +03:00
Marek Vavruša
95d926ec14 AggregateFunctionTopK: smaller initial table size
By default start with 2^4 elements
2017-05-13 03:58:05 +04:00
Marek Vavruša
45bd332460 AggregateFunctionTopK: fix memory usage, performance
* allow separate table key / hash key, and use
  std::string / StringRef for generic variant as
  it has built-in storage and StringRef is supported
  by the hash table, this avoids infinitely
  growing arena with serialised keys
* use power-of-2 size for alpha vector for faster
  binning without using modulo
* use custom grower and allocator for SpaceSaving
  to start with smaller tables
* store computed hash in counter for faster
  reinsertion of smallest element
2017-05-11 18:52:49 +04:00
Alexey Milovidov
d3e6321967 AggregateFunctionTopK: minor modifications [#CLICKHOUSE-2]. 2017-05-05 16:36:02 -07:00
Marek Vavruša
5f1e65b252 AggregateFunctions: implemented topK(n)
This implements a new function for approximate
computation of the most frequent entries using
Filtered Space Saving with a merge step adapted
from Parallel Space Saving paper.

It works better for cases where GROUP BY x
is impractical due to high cardinality of x,
such as top IP addresses or top search queries.
2017-05-03 23:09:52 -07:00