bool CDictionary::IsExist(char *sWord, int nHandle)
{
char sWordFind[WORD_MAXLENGTH-2];
int nPos;
if(!PreProcessing(sWord, &nPos,sWordFind))
return false;
return(FindInOriginalTable(nPos,sWordFind,nHandle)||FindInModifyTable(nPos,sWordFind,nHandle));
}
bool CDictionary::DelItem(char *sWord,int nHandle)
{
char sWordDel[WORD_MAXLENGTH-2];
int nPos,nFoundPos,nTemp;
PWORD_CHAIN pPre,pTemp,pCur;
if(!PreProcessing(sWord, &nPos,sWordDel))
return false;
if (nPos < 0)
return false;
if(FindInOriginalTable(nPos,sWordDel,nHandle,&nFoundPos))
{
if(!m_pModifyTable)//Not prepare the buffer
{
m_pModifyTable=new MODIFY_TABLE[CC_NUM];
memset(m_pModifyTable,0,CC_NUM*sizeof(MODIFY_TABLE));
}
m_IndexTable[nPos].pWordItemHead[nFoundPos].nFrequency=-1;
m_pModifyTable[nPos].nDelete+=1;
if(nHandle==-1)//Remove all items which word is sWordDel,ignoring the handle
{
/* nTemp=nFoundPos-1;//Check its previous position
while(nTemp>0&&strcmp(m_IndexTable[nPos].pWordItemHead[nFoundPos].sWord,sWordDel)==0)
{
m_IndexTable[nPos].pWordItemHead[nTemp].nFrequency=-1;
m_pModifyTable[nPos].nDelete+=1;
nTemp-=1;
}
*/ nTemp=nFoundPos+1;//Check its previous position
while(nTemp<m_IndexTable[nPos].nCount&&strcmp(m_IndexTable[nPos].pWordItemHead[nFoundPos].sWord,sWordDel)==0)
{
m_IndexTable[nPos].pWordItemHead[nTemp].nFrequency=-1;
m_pModifyTable[nPos].nDelete+=1;
nTemp+=1;
}
}
return true;
}
//Operation in the modify table and its items
if(FindInModifyTable(nPos,sWordDel,nHandle,&pPre))
{
pCur=m_pModifyTable[nPos].pWordItemHead;
if(pPre!=NULL)
pCur=pPre->next;
while(pCur!=NULL && _stricmp(pCur->data.sWord,sWordDel)==0&&(pCur->data.nHandle==nHandle||nHandle<0))
{
pTemp=pCur;
if(pPre!=NULL)//pCur is the first item
pPre->next=pCur->next;
else
m_pModifyTable[nPos].pWordItemHead=pCur->next;
pCur=pCur->next;
delete pTemp->data.sWord;//Delete the word
delete pTemp;
}
return true;
}
return false;
}
bool CDictionary::GetHandle(char *sWord,int *pnCount,int *pnHandle,int *pnFrequency)
{
char sWordGet[WORD_MAXLENGTH-2];
int nPos,nFoundPos,nTemp;
PWORD_CHAIN pPre,pCur;
*pnCount=0;
if(!PreProcessing(sWord, &nPos,sWordGet))
return false;
if (nPos < 0)
return false;
if(FindInOriginalTable(nPos,sWordGet,-1,&nFoundPos))
{
pnHandle[*pnCount]=m_IndexTable[nPos].pWordItemHead[nFoundPos].nHandle;
pnFrequency[*pnCount]=m_IndexTable[nPos].pWordItemHead[nFoundPos].nFrequency;
*pnCount+=1;
/* nTemp=nFoundPos-1;//Check its previous position
while(nTemp>0&&strcmp(m_IndexTable[nPos].pWordItemHead[nTemp].sWord,sWordGet)==0)
{
pnHandle[*pnCount]=m_IndexTable[nPos].pWordItemHead[nTemp].nHandle;
pnFrequency[*pnCount]=m_IndexTable[nPos].pWordItemHead[nTemp].nFrequency;
*pnCount+=1;
nTemp-=1;
}
*/ nTemp=nFoundPos+1;//Check its previous position
while(nTemp<m_IndexTable[nPos].nCount&&strcmp(m_IndexTable[nPos].pWordItemHead[nTemp].sWord,sWordGet)==0)
{
pnHandle[*pnCount]=m_IndexTable[nPos].pWordItemHead[nTemp].nHandle;
pnFrequency[*pnCount]=m_IndexTable[nPos].pWordItemHead[nTemp].nFrequency;
*pnCount+=1;
nTemp+=1;
}
return true;
}
//Operation in the index table and its items
if(FindInModifyTable(nPos,sWordGet,-1,&pPre))
{
pCur=m_pModifyTable[nPos].pWordItemHead;
if(pPre!=NULL)
pCur=pPre->next;
while(pCur!=NULL && _stricmp(pCur->data.sWord,sWordGet)==0)
{
pnHandle[*pnCount]=pCur->data.nHandle;
pnFrequency[*pnCount]=pCur->data.nFrequency;
*pnCount+=1;
pCur=pCur->next;
}
return true;
}
return false;
}
int CDictionary::GetFrequency(char *sWord, int nHandle)
{
char sWordFind[WORD_MAXLENGTH-2];
int nPos,nIndex;
PWORD_CHAIN pFound;
if(!PreProcessing(sWord, &nPos,sWordFind))
return 0;
if(FindInOriginalTable(nPos,sWordFind,nHandle,&nIndex))
{
return m_IndexTable[nPos].pWordItemHead[nIndex].nFrequency;
}
if(FindInModifyTable(nPos,sWordFind,nHandle,&pFound))
{
return pFound->data.nFrequency;
}
return 0;
}
/*********************************************************************
*
* Func Name : GetMaxMatch
*
* Description: Get the max match to the word
*
*
* Parameters : nHandle: the only handle which will be attached to the word
* Returns : success or fail
* Author : Kevin Zhang
* History :
* 1.create 2002-1-21
*********************************************************************/
bool CDictionary::GetMaxMatch(char *sWord, char *sWordRet,int *npHandleRet)
{
char sWordGet[WORD_MAXLENGTH-2],sFirstChar[3];
int nPos,nFoundPos,nTemp;
PWORD_CHAIN pCur;
*npHandleRet=-1;
if(!PreProcessing(sWord, &nPos,sWordGet))
return false;
if (nPos < 0) return false;
sWordRet[0]=0;
strncpy(sFirstChar,sWord,strlen(sWord)-strlen(sWordGet));//Get the first char
sFirstChar[strlen(sWord)-strlen(sWordGet)]=0;//Set the end flag
FindInOriginalTable(nPos,sWordGet,-1,&nFoundPos);
nTemp=nFoundPos;//Check its previous position
if(nFoundPos==-1)
nTemp=0;
while(nTemp<m_IndexTable[nPos].nCount&&CC_Find(m_IndexTable[nPos].pWordItemHead[nTemp].sWord,sWordGet)!=m_IndexTable[nPos].pWordItemHead[nTemp].sWord)
{//Get the next
nTemp+=1;
}
if(nTemp<m_IndexTable[nPos].nCount&&CC_Find(m_IndexTable[nPos].pWordItemHead[nTemp].sWord,sWordGet)==m_IndexTable[nPos].pWordItemHead[nTemp].sWord)
{
strcpy(sWordRet,sFirstChar);
strcat(sWordRet,m_IndexTable[nPos].pWordItemHead[nTemp].sWord);
*npHandleRet=m_IndexTable[nPos].pWordItemHead[nTemp].nHandle;
return true;
}//Cannot get the item and retrieve the modified data if exists
//Operation in the index table and its items
if(m_pModifyTable&&m_pModifyTable[nPos].pWordItemHead)//Exists
pCur=m_pModifyTable[nPos].pWordItemHead;
else
pCur=NULL;
while(pCur!=NULL&&strcmp(pCur->data.sWord,sWordGet)<=0&&CC_Find(pCur->data.sWord,sWordGet)!=pCur->data.sWord)//
{
pCur=pCur->next;
}
if(pCur!=NULL&&CC_Find(pCur->data.sWord,sWordGet)!=pCur->data.sWord)
{//Get it
strcpy(sWordRet,sFirstChar);
strcat(sWordRet,pCur->data.sWord);
*npHandleRet=pCur->data.nHandle;
return true;
}
return false;
}
bool CDictionary::AddItem(char *sWord, int nHandle,int nFrequency)
{
char sWordAdd[WORD_MAXLENGTH-2];
int nPos,nFoundPos;
PWORD_CHAIN pRet,pTemp,pNext;
int i=0;
if(!PreProcessing(sWord, &nPos,sWordAdd,true))
return false;
if(FindInOriginalTable(nPos,sWordAdd,nHandle,&nFoundPos))
{//The word exists in the original table, so add the frequency
//Operation in the index table and its items
if(m_IndexTable[nPos].pWordItemHead[nFoundPos].nFrequency==-1)
{//The word item has been removed
m_IndexTable[nPos].pWordItemHead[nFoundPos].nFrequency=nFrequency;
if(!m_pModifyTable)//Not prepare the buffer
{
m_pModifyTable=new MODIFY_TABLE[CC_NUM];
memset(m_pModifyTable,0,CC_NUM*sizeof(MODIFY_TABLE));
}
m_pModifyTable[nPos].nDelete-=1;
}
else
m_IndexTable[nPos].pWordItemHead[nFoundPos].nFrequency+=nFrequency;
return true;
}
//The items not exists in the index table.
//As following, we have to find the item whether exists in the modify data region
//If exists, change the frequency .or else add a item
if(!m_pModifyTable)//Not prepare the buffer
{
m_pModifyTable=new MODIFY_TABLE[CC_NUM];
memset(m_pModifyTable,0,CC_NUM*sizeof(MODIFY_TABLE));
}
if(FindInModifyTable(nPos,sWordAdd,nHandle,&pRet))
{
if(pRet!=NULL)
pRet=pRet->next;
else
pRet=m_pModifyTable[nPos].pWordItemHead;
pRet->data.nFrequency+=nFrequency;
return true;
}
//find the proper position to add the word to the modify data table and link
pTemp=new WORD_CHAIN;//Allocate the word chain node
if(pTemp==NULL)//Allocate memory failure
return false;
memset(pTemp,0,sizeof(WORD_CHAIN));//init it with 0
pTemp->data.nHandle=nHandle;//store the handle
pTemp->data.nWordLen=strlen(sWordAdd);
pTemp->data.sWord=new char[1+pTemp->data.nWordLen];
strcpy(pTemp->data.sWord,sWordAdd);
pTemp->data.nFrequency=nFrequency;
pTemp->next=NULL;
if(pRet!=NULL)
{
pNext=pRet->next;//Get the next item before the current item
pRet->next=pTemp;//link the node to the chain
}
else
{
pNext=m_pModifyTable[nPos].pWordItemHead;
m_pModifyTable[nPos].pWordItemHead=pTemp;//Set the pAdd as the head node
}
pTemp->next=pNext;//Very important!!!! or else it will lose some node
//Modify in 2001-10-29
m_pModifyTable[nPos].nCount++;//the number increase by one
return true;
}
int main(int argc, char* argv[])
{
char lib1FileName[1000], lib2FileName[1000], chromFileName[1000],configFileName[2000], projectName[1000];
CONFIG_STRUCT config;
CHROM_INFO chroms[MAX_CHROM_NUM];
int chromNum;
double noiseRate, l1Ratio, l2Ratio;
int maxL1Count, maxL2Count;
int i;
if (argc!=6)
{
printf("Usage: <library 1 tag file name> <library 2 tag file name> <chromosome length file name> <config file name> <project name>\n");
return 0;
}
strcpy(lib1FileName, argv[1]);
strcpy(lib2FileName, argv[2]);
strcpy(chromFileName, argv[3]);
strcpy(configFileName, argv[4]);
strcpy(projectName, argv[5]);
chromNum = GetChromInfo(chromFileName, chroms);
if (chromNum<=0)
{
return 0;
}
printf("chromosome length information read. chromNum = %d!\n", chromNum);
for (i=0;i<chromNum;i++)
{
chroms[i].l1PosTagNum = 0;
chroms[i].l2PosTagNum = 0;
chroms[i].l1NegTagNum = 0;
chroms[i].l2NegTagNum = 0;
chroms[i].l1PeakNum = 0;
chroms[i].l2PeakNum = 0;
}
if (ReadConfigFile(configFileName, &config)<0)
{
printf("config file not complet. Default setting used.\n");
}
printf("config file read.\n");
printf("fragmentSize = %d\n", config.fragmentSize);
printf("isStrandSensitiveMode = %d\n", config.isStrandSensitiveMode);
printf("slidingWinSize = %d\n", config.slidingWinSize);
printf("movingStep = %d\n", config.movingStep);
printf("outputNum = %d\n", config.outputNum);
printf("minCount = %d\n", config.minCount);
printf("minScore = %f\n", config.minScore);
printf("bootstrapPass = %d\n", config.bootstrapPass);
printf("randSeed = %d\n", config.randomSeed);
printf("reading tag files......\n");
if (!GetTags(lib1FileName, lib2FileName, chroms, chromNum))
{
printf("tag file error!\n");
return 0;
}
printf("tag file read.\n");
printf("pre-processing......\n");
if (PreProcessing(chroms, chromNum)<0)
{
printf("Preprocessing failed!\n CCAT aborted!\n");
return -1;
}
printf("pre-processing finished.\n");
printf("estimating noise rate......\n");
srand(config.randomSeed);
noiseRate = ComputeNoiseRate(chroms,chromNum, &l1Ratio,&l2Ratio, config);
printf("noise rate = %f\n", noiseRate);
printf("peak-finding......\n");
if (PeakFinding(chroms,chromNum,l1Ratio,l2Ratio, &maxL1Count, &maxL2Count, config)<=0)
{
return -1;
}
printf("peak-finding finished.\n");
printf("Significance Analysis......\n");
if (SignificanceAnalysis(chroms, chromNum, l1Ratio, l2Ratio, maxL1Count, maxL2Count, config)<0)
{
printf("DPS paramter esitmation failed!\n");
return -1;
}
printf("Significance analysis finished.\n");
printf("saving results.......\n");
OutputFiles(chroms, chromNum, projectName, config);
printf("results saved.\n");
FreeChromMem(chroms,chromNum);
printf("CCAT process completed!\n");
return 0;
}
void KernelBand<NLSSystemObject>::NumericalJacobian(const Eigen::VectorXd& x, const Eigen::VectorXd& f, Eigen::VectorXd& x_dimensions, const bool max_constraints, const Eigen::VectorXd& xMax)
{
if (pre_processing_ == false)
PreProcessing();
const double tstart = OpenSMOKE::OpenSMOKEGetCpuTime();
const double ZERO_DER = 1.e-8;
const double ETA2 = std::sqrt(OpenSMOKE::OPENSMOKE_MACH_EPS_DOUBLE);
// Save the original vector
x_plus_ = x;
// Loop
for (int group = 1; group <= ngroups_; group++)
{
if (max_constraints == false)
{
for (int j = group - 1; j < static_cast<int>(this->ne_); j += width_)
{
const double xh = std::fabs(x(j));
const double xdh = std::fabs(x_dimensions(j));
hJ_(j) = ETA2*std::max(xh, xdh);
hJ_(j) = std::max(hJ_(j), ZERO_DER);
hJ_(j) = std::min(hJ_(j), 0.001 + 0.001*std::fabs(xh));
x_plus_(j) += hJ_(j);
}
}
else
{
for (int j = group - 1; j < static_cast<int>(this->ne_); j += width_)
{
const double xh = std::fabs(x(j));
const double xdh = std::fabs(x_dimensions(j));
hJ_(j) = ETA2*std::max(xh, xdh);
hJ_(j) = std::max(hJ_(j), ZERO_DER);
hJ_(j) = std::min(hJ_(j), 0.001 + 0.001*std::fabs(xh));
if (xh + hJ_(j) > xMax(j))
hJ_(j) = -hJ_(j);
x_plus_(j) += hJ_(j);
}
}
this->Equations(x_plus_, f_plus_);
for (int j = group - 1; j < static_cast<int>(this->ne_); j += width_)
{
x_plus_(j) = x(j);
double* col_j = BAND_COL(J_, j);
int i1 = std::max(0, j - J_->nUpper());
int i2 = std::min(j + J_->nLower(), static_cast<int>(this->ne_ - 1));
for (int i = i1; i <= i2; i++)
BAND_COL_ELEM(col_j, i, j) = (f_plus_(i) - f(i)) / hJ_(j);
}
}
const double tend = OpenSMOKE::OpenSMOKEGetCpuTime();
numberOfSystemCallsForJacobian_ += this->ngroups_;
numberOfJacobianFullAssembling_++;
cpuTimeSingleJacobianFullAssembling_ = tend - tstart;
cpuTimeJacobianFullAssembling_ += cpuTimeSingleJacobianFullAssembling_;
}
