/*
* processFile should:
* 1) Create a new instance of your MyTrends class
* 2) Read in the file whose words you want to count: The file name is passed in as fname
* 3) Add all words to the Trends data structure, and calculate the time per call to addtoTrends
* 4) Use getNthPopular and getCount to print out the total results in fname + ".out"
*/
int processFile(const char* fname){
Trends* tr = new MyTrends();
// First, read in the file
ifstream in(fname);
string s;
vector<string> wordlist;
while(in >> s){
// s = reduceWord(s);
if (s.size() > 0)
wordlist.push_back(s);
}
// We only want to time how long addToTrends takes, so we get
// the starting time, which is the clock time, in milliseconds
std::clock_t start, end;
start = std::clock();
// int start = timeGetTime();
// Now add all the words to the Trends data structure
for(unsigned int i = 0; i < wordlist.size(); i++){
tr->increaseCount(wordlist[i], 1);
}
// Now get the end time
// int end = timeGetTime();
end = std::clock();
double millisec = (end - start) / ((double)CLOCKS_PER_SEC / 1000);
std::cout << "Time: " << ((1000.0 * millisec) / wordlist.size()) << " microseconds per word" << endl;
// Now we will print out the complete results. This could be REALLY slow, if
// your getNthPopular is not a little bit smart.
int p = ((string)fname).rfind('.');
string outfname = ((string)fname).substr(0, p) + ".out";
ofstream out(outfname.c_str());
for(int i=0; i< tr->numEntries(); i++){
string s = tr->getNthPopular(i);
out << tr->getCount(s) << ": " << s << endl;
}
out.close();
delete tr;
return end - start;
}
/**
* This tests a simple (but unlikely) use case, which is to read in all the data, and then print out the data in sorted order
* based on popularity.
*
* Compare your 28885.txt.out to 28885_txt.out, using diff,s to see if your code is producing correct output.
*/
double useCase_addAllThenGetInOrder(){
Trends* tr = new smarterTrends(); //You will need to change this to match your own class!
std::vector<std::string> wordlist = getWordList("data/28885.txt");
//We only want to time how long addToTrends takes, so we get
// the starting time, which is the clock time, in milliseconds
double start = getTimeInMillis();
//Now add all the words to the Trends data structure
for(unsigned int i=0; i<wordlist.size(); i++){
tr->increaseCount(wordlist[i],1);
}
//Now get the end time
double end = getTimeInMillis();
std::cout << "increaseCount time: " << (end-start)/wordlist.size() << " ms per word" << std::endl;
//Now we will print out the complete results. This could be REALLY clow, if
// your getNthPopular is not a little bit smart.
std::string outfname = "data/28885.txt.out";
std::ofstream out(outfname.c_str());
start = getTimeInMillis();
for(unsigned int i=0; i< tr->numEntries(); i++){
std::string s = tr->getNthPopular(i);
out << tr->getCount(s) << ": " << s << std::endl;
}
out.close();
end = getTimeInMillis();
std::cout << "getNth followed by getCount, time: " << (end - start) / tr->numEntries() << " ms per entry" << std::endl;
delete tr;
return end - start;
}
/*
* processFile should:
* 1) Create a new instance of your trends class
* 2) Read in the file whose words you want to count: The file name is passed in as fname
* 3) Add all words to the Trends data structure, and calculate the time per call to addtoTrends
* 4) Use getNthPopular and getCount to print out the total results in fname + ".out"
*/
int processFile(const char* fname){
Trends* tr = new brinkmwjTrends(); //You will need to change this to match your own class!
//First, read in the file
// I'm not doing much error checking here, because I am a bad person
std::ifstream in(fname);
std::string s;
std::vector<std::string> wordlist;
while(in >> s){
wordlist.push_back(s);
}
//We only want to time how long addToTrends takes, so we get
// the starting time, which is the clock time, in milliseconds
clock_t start = clock();
//Now add all the words to the Trends data structure
for(unsigned int i=0; i<wordlist.size(); i++){
tr->increaseCount(wordlist[i],1);
}
//Now get the end time
clock_t end = clock();
std::cout << "Time: " << ((1000.0*(end - start)/CLOCKS_PER_SEC)/wordlist.size()) << " ms per word" << std::endl;
//Now we will print out the complete results. This could be REALLY clow, if
// your getNthPopular is not a little bit smart.
std::string outfname = fname;
outfname = outfname + ".out";
std::ofstream out(outfname.c_str());
for(int i=0; i< tr->numEntries(); i++){
std::string s = tr->getNthPopular(i);
out << tr->getCount(s) << ": " << s << std::endl;
}
out.close();
delete tr;
return end - start;
}
