database/mariadb-columnstore-engine
1
0
Fork 0
mirror of https://github.com/mariadb-corporation/mariadb-columnstore-engine.git synced 2025-07-30 19:23:07 +03:00
Code Activity

MCOL-521 add regr_intecept and regr_r2

Browse Source
This commit is contained in:
David Hall
2018-09-28 13:51:43 -05:00
parent b8bf311c51
commit dc9ba90f96
11 changed files with 1058 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

216
utils/regr/regr_r2.cpp Normal file
View File

@ -0,0 +1,216 @@
/* Copyright (C) 2017 MariaDB Corporaton
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License
as published by the Free Software Foundation; version 2 of
the License.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
MA 02110-1301, USA. */
#include <sstream>
#include <cstring>
#include <typeinfo>
#include "regr_r2.h"
#include "bytestream.h"
#include "objectreader.h"
using namespace mcsv1sdk;
class Add_regr_r2_ToUDAFMap
{
public:
Add_regr_r2_ToUDAFMap()
{
UDAFMap::getMap()["regr_r2"] = new regr_r2();
}
};
static Add_regr_r2_ToUDAFMap addToMap;
// Use the simple data model
struct regr_r2_data
{
uint64_t cnt;
double sumx;
double sumx2; // sum of (x squared)
double sumy;
double sumy2; // sum of (y squared)
double sumxy; // sum of x * y
};
mcsv1_UDAF::ReturnCode regr_r2::init(mcsv1Context* context,
ColumnDatum* colTypes)
{
if (context->getParameterCount() != 2)
{
// The error message will be prepended with
// "The storage engine for the table doesn't support "
context->setErrorMessage("regr_r2() with other than 2 arguments");
return mcsv1_UDAF::ERROR;
}
context->setUserDataSize(sizeof(regr_r2_data));
context->setResultType(CalpontSystemCatalog::DOUBLE);
context->setColWidth(8);
context->setScale(colTypes[0].scale + 8);
context->setPrecision(19);
context->setRunFlag(mcsv1sdk::UDAF_IGNORE_NULLS);
return mcsv1_UDAF::SUCCESS;
}
mcsv1_UDAF::ReturnCode regr_r2::reset(mcsv1Context* context)
{
struct regr_r2_data* data = (struct regr_r2_data*)context->getUserData()->data;
data->cnt = 0;
data->sumx = 0.0;
data->sumx2 = 0.0;
data->sumy = 0.0;
data->sumy2 = 0.0;
data->sumxy = 0.0;
return mcsv1_UDAF::SUCCESS;
}
mcsv1_UDAF::ReturnCode regr_r2::nextValue(mcsv1Context* context, ColumnDatum* valsIn)
{
static_any::any& valIn_y = valsIn[0].columnData;
static_any::any& valIn_x = valsIn[1].columnData;
struct regr_r2_data* data = (struct regr_r2_data*)context->getUserData()->data;
double valx = 0.0;
double valy = 0.0;
valx = convertAnyTo<double>(valIn_x);
valy = convertAnyTo<double>(valIn_y);
// For decimal types, we need to move the decimal point.
uint32_t scaley = valsIn[0].scale;
if (valy != 0 && scaley > 0)
{
valy /= pow(10.0, (double)scaley);
}
data->sumy += valy;
data->sumy2 += valy*valy;
// For decimal types, we need to move the decimal point.
uint32_t scalex = valsIn[1].scale;
if (valx != 0 && scalex > 0)
{
valx /= pow(10.0, (double)scaley);
}
data->sumx += valx;
data->sumx2 += valx*valx;
data->sumxy += valx*valy;
++data->cnt;
return mcsv1_UDAF::SUCCESS;
}
mcsv1_UDAF::ReturnCode regr_r2::subEvaluate(mcsv1Context* context, const UserData* userDataIn)
{
if (!userDataIn)
{
return mcsv1_UDAF::SUCCESS;
}
struct regr_r2_data* outData = (struct regr_r2_data*)context->getUserData()->data;
struct regr_r2_data* inData = (struct regr_r2_data*)userDataIn->data;
outData->sumx += inData->sumx;
outData->sumx2 += inData->sumx2;
outData->sumy += inData->sumy;
outData->sumy2 += inData->sumy2;
outData->sumxy += inData->sumxy;
outData->cnt += inData->cnt;
return mcsv1_UDAF::SUCCESS;
}
mcsv1_UDAF::ReturnCode regr_r2::evaluate(mcsv1Context* context, static_any::any& valOut)
{
struct regr_r2_data* data = (struct regr_r2_data*)context->getUserData()->data;
double N = data->cnt;
if (N > 0)
{
double sumx = data->sumx;
double sumy = data->sumy;
double sumx2 = data->sumx2;
double sumy2 = data->sumy2;
double sumxy = data->sumxy;
double var_popx = (sumx2 - (sumx * sumx / N)) / N;
if (var_popx == 0)
{
// When var_popx is 0, NULL is the result.
return mcsv1_UDAF::SUCCESS;
}
double var_popy = (sumy2 - (sumy * sumy / N)) / N;
if (var_popy == 0)
{
// When var_popy is 0, 1 is the result
valOut = 1.0;
return mcsv1_UDAF::SUCCESS;
}
double std_popx = sqrt(var_popx);
double std_popy = sqrt(var_popy);
double covar_pop = (sumxy - ((sumx * sumy) / N)) / N;
double corr = covar_pop / (std_popy * std_popx);
valOut = corr * corr;
}
return mcsv1_UDAF::SUCCESS;
}
mcsv1_UDAF::ReturnCode regr_r2::dropValue(mcsv1Context* context, ColumnDatum* valsDropped)
{
static_any::any& valIn_y = valsDropped[0].columnData;
static_any::any& valIn_x = valsDropped[1].columnData;
struct regr_r2_data* data = (struct regr_r2_data*)context->getUserData()->data;
double valx = 0.0;
double valy = 0.0;
valx = convertAnyTo<double>(valIn_x);
valy = convertAnyTo<double>(valIn_y);
// For decimal types, we need to move the decimal point.
uint32_t scaley = valsDropped[0].scale;
if (valy != 0 && scaley > 0)
{
valy /= pow(10.0, (double)scaley);
}
data->sumy -= valy;
data->sumy2 -= valy*valy;
// For decimal types, we need to move the decimal point.
uint32_t scalex = valsDropped[1].scale;
if (valx != 0 && scalex > 0)
{
valx /= pow(10.0, (double)scaley);
}
data->sumx -= valx;
data->sumx2 -= valx*valx;
data->sumxy -= valx*valy;
--data->cnt;
return mcsv1_UDAF::SUCCESS;
}