rmontanana/Platform
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Convert DatasetsConsole & ResultsDatasetConsole to string output

Browse Source
This commit is contained in:
Ricardo Montañana Gómez
2024-03-16 13:48:49 +01:00
parent b9af086c29
commit fa366a4c22
9 changed files with 201 additions and 180 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
src/CMakeLists.txt
View File

@@ -40,8 +40,8 @@ target_link_libraries(b_grid ${MPI_CXX_LIBRARIES} "${PyClassifiers}" "${BayesNet
add_executable(b_list commands/b_list.cpp ${list_sources}
common/Datasets.cpp common/Dataset.cpp
main/Models.cpp
reports/ReportExcel.cpp reports/ExcelFile.cpp reports/ReportBase.cpp reports/DatasetsExcel.cpp
results/Result.cpp results/ResultsDatasetExcel.cpp results/ResultsDataset.cpp
reports/ReportExcel.cpp reports/ExcelFile.cpp reports/ReportBase.cpp reports/DatasetsExcel.cpp reports/DatasetsConsole.cpp
results/Result.cpp results/ResultsDatasetExcel.cpp results/ResultsDataset.cpp results/ResultsDatasetConsole.cpp
)
target_link_libraries(b_list "${PyClassifiers}" "${BayesNet}" ArffFiles mdlp ${Python3_LIBRARIES} "${TORCH_LIBRARIES}" ${LIBTORCH_PYTHON} Boost::python Boost::numpy "${XLSXWRITER_LIB}")
@@ -61,7 +61,7 @@ list(TRANSFORM manage_sources PREPEND manage/)
add_executable(
b_manage commands/b_manage.cpp ${manage_sources}
common/Datasets.cpp common/Dataset.cpp
reports/ReportConsole.cpp reports/ReportExcel.cpp reports/ReportExcelCompared.cpp reports/ReportBase.cpp reports/ExcelFile.cpp
results/Result.cpp
reports/ReportConsole.cpp reports/ReportExcel.cpp reports/ReportExcelCompared.cpp reports/ReportBase.cpp reports/ExcelFile.cpp reports/DatasetsConsole.cpp
results/Result.cpp results/ResultsDatasetConsole.cpp results/ResultsDataset.cpp
)
target_link_libraries(b_manage "${TORCH_LIBRARIES}" "${XLSXWRITER_LIB}" ArffFiles mdlp)

4
src/commands/b_list.cpp
View File

@@ -9,10 +9,10 @@
#include "common/Colors.h"
#include "common/Datasets.h"
#include "reports/DatasetsExcel.h"
#include "reports/DatasetsConsole.hpp"
#include "reports/DatasetsConsole.h"
#include "results/ResultsDataset.h"
#include "results/ResultsDatasetExcel.h"
#include "results/ResultsDatasetConsole.hpp"
#include "results/ResultsDatasetConsole.h"
#include "config.h"

2
src/manage/ManageResults.cpp
View File

@@ -5,7 +5,7 @@
#include "common/Paths.h"
#include "CommandParser.h"
#include "ManageResults.h"
// #include "reports/DatasetsConsole.hpp"
//#include "reports/DatasetsConsole.h"
#include "reports/ReportConsole.h"
#include "reports/ReportExcel.h"
#include "reports/ReportExcelCompared.h"

56
src/reports/DatasetsConsole.cpp Normal file
View File

@@ -0,0 +1,56 @@
#include "common/Colors.h"
#include "common/Datasets.h"
#include "common/Paths.h"
#include "DatasetsConsole.h"
namespace platform {
const int DatasetsConsole::BALANCE_LENGTH = 75;
std::string DatasetsConsole::outputBalance(const std::string& balance)
{
auto temp = std::string(balance);
while (temp.size() > DatasetsConsole::BALANCE_LENGTH - 1) {
auto part = temp.substr(0, DatasetsConsole::BALANCE_LENGTH);
output << part << std::endl;
output << setw(52) << " ";
temp = temp.substr(DatasetsConsole::BALANCE_LENGTH);
}
return temp;
}
void DatasetsConsole::list_datasets()
{
auto datasets = platform::Datasets(false, platform::Paths::datasets());
locale mylocale(std::cout.getloc(), new separated_datasets);
locale::global(mylocale);
output.imbue(mylocale);
output << Colors::GREEN() << " # Dataset Sampl. Feat. Cls Balance" << std::endl;
std::string balanceBars = std::string(DatasetsConsole::BALANCE_LENGTH, '=');
output << "=== ============================== ====== ===== === " << balanceBars << std::endl;
int num = 0;
for (const auto& dataset : datasets.getNames()) {
auto color = num % 2 ? Colors::CYAN() : Colors::BLUE();
output << color << setw(3) << right << num++ << " ";
output << setw(30) << left << dataset << " ";
datasets.loadDataset(dataset);
auto nSamples = datasets.getNSamples(dataset);
output << setw(6) << right << nSamples << " ";
output << setw(5) << right << datasets.getFeatures(dataset).size() << " ";
output << setw(3) << right << datasets.getNClasses(dataset) << " ";
std::stringstream oss;
std::string sep = "";
for (auto number : datasets.getClassesCounts(dataset)) {
oss << sep << std::setprecision(2) << fixed << (float)number / nSamples * 100.0 << "% (" << number << ")";
sep = " / ";
}
auto balance = outputBalance(oss.str());
output << balance << std::endl;
// Store data for Excel report
data[dataset] = json::object();
data[dataset]["samples"] = nSamples;
data[dataset]["features"] = datasets.getFeatures(dataset).size();
data[dataset]["classes"] = datasets.getNClasses(dataset);
data[dataset]["balance"] = oss.str();
}
numLines = num + 2;
}
}

32
src/reports/DatasetsConsole.h Normal file
View File

@@ -0,0 +1,32 @@
#pragma once
#include <locale>
#include <sstream>
#include <nlohmann/json.hpp>
namespace platform {
using json = nlohmann::json;
struct separated_datasets : numpunct<char> {
char do_decimal_point() const { return ','; }
char do_thousands_sep() const { return '.'; }
std::string do_grouping() const { return "\03"; }
};
class DatasetsConsole {
public:
static const int BALANCE_LENGTH;
DatasetsConsole() = default;
~DatasetsConsole() = default;
std::string getOutput() const { return output.str(); }
int getNumLines() const { return numLines; }
json& getData() { return data; }
std::string outputBalance(const std::string& balance);
void list_datasets();
private:
std::stringstream output;
json data;
int numLines = 0;
};
}

78
src/reports/DatasetsConsole.hpp
View File

@@ -1,78 +0,0 @@
#pragma once
#include <locale>
#include <map>
#include <sstream>
#include <nlohmann/json.hpp>
#include "common/Colors.h"
#include "common/Datasets.h"
namespace platform {
const int BALANCE_LENGTH = 75;
struct separated_datasets : numpunct<char> {
char do_decimal_point() const { return ','; }
char do_thousands_sep() const { return '.'; }
std::string do_grouping() const { return "\03"; }
};
class DatasetsConsole {
public:
DatasetsConsole() = default;
~DatasetsConsole() = default;
std::string getOutput() const { return output.str(); }
int getNumLines() const { return numLines; }
json& getData() { return data; }
std::string outputBalance(const std::string& balance)
{
auto temp = std::string(balance);
while (temp.size() > BALANCE_LENGTH - 1) {
auto part = temp.substr(0, BALANCE_LENGTH);
output << part << std::endl;
output << setw(52) << " ";
temp = temp.substr(BALANCE_LENGTH);
}
return temp;
}
void list_datasets()
{
auto datasets = platform::Datasets(false, platform::Paths::datasets());
locale mylocale(std::cout.getloc(), new separated_datasets);
locale::global(mylocale);
output.imbue(mylocale);
output << Colors::GREEN() << " # Dataset Sampl. Feat. Cls Balance" << std::endl;
std::string balanceBars = std::string(BALANCE_LENGTH, '=');
output << "=== ============================== ====== ===== === " << balanceBars << std::endl;
int num = 0;
for (const auto& dataset : datasets.getNames()) {
auto color = num % 2 ? Colors::CYAN() : Colors::BLUE();
output << color << setw(3) << right << num++ << " ";
output << setw(30) << left << dataset << " ";
datasets.loadDataset(dataset);
auto nSamples = datasets.getNSamples(dataset);
output << setw(6) << right << nSamples << " ";
output << setw(5) << right << datasets.getFeatures(dataset).size() << " ";
output << setw(3) << right << datasets.getNClasses(dataset) << " ";
std::stringstream oss;
std::string sep = "";
for (auto number : datasets.getClassesCounts(dataset)) {
oss << sep << std::setprecision(2) << fixed << (float)number / nSamples * 100.0 << "% (" << number << ")";
sep = " / ";
}
auto balance = outputBalance(oss.str());
output << balance << std::endl;
// Store data for Excel report
data[dataset] = json::object();
data[dataset]["samples"] = nSamples;
data[dataset]["features"] = datasets.getFeatures(dataset).size();
data[dataset]["classes"] = datasets.getNClasses(dataset);
data[dataset]["balance"] = oss.str();
}
numLines = num + 2;
}
private:
std::stringstream output;
json data;
int numLines = 0;
};
}

79
src/results/ResultsDatasetConsole.cpp Normal file
View File

@@ -0,0 +1,79 @@
#include "common/Colors.h"
#include "results/ResultsDataset.h"
#include "ResultsDatasetConsole.h"
namespace platform {
void ResultsDatasetsConsole::list_results(const std::string& dataset, const std::string& score, const std::string& model)
{
output.str("");
auto results = platform::ResultsDataset(dataset, model, score);
results.load();
results.sortModel();
if (results.empty()) {
output << Colors::RED() << "No results found for dataset " << dataset << " and model " << model << Colors::RESET() << std::endl;
return;
}
int maxModel = results.maxModelSize();
int maxHyper = results.maxHyperSize();
double maxResult = results.maxResultScore();
// Build data for the Report
json data = json::object();
data["results"] = json::array();
data["max_models"] = json::object(); // Max score per model
for (const auto& result : results) {
auto results = result.getData();
if (!data["max_models"].contains(result.getModel())) {
data["max_models"][result.getModel()] = 0;
}
for (const auto& item : results["results"]) {
if (item["dataset"] == dataset) {
// Store data for Excel report
json res = json::object();
res["date"] = result.getDate();
res["time"] = result.getTime();
res["model"] = result.getModel();
res["score"] = item["score"].get<double>();
res["hyperparameters"] = item["hyperparameters"].dump();
data["results"].push_back(res);
if (item["score"].get<double>() > data["max_models"][result.getModel()]) {
data["max_models"][result.getModel()] = item["score"].get<double>();
}
break;
}
}
}
//
// List the results
//
output << Colors::GREEN() << "Results of dataset " << dataset << " - for " << model << " model" << std::endl;
output << "There are " << results.size() << " results" << std::endl;
output << Colors::GREEN() << " # " << std::setw(maxModel + 1) << std::left << "Model" << "Date Time Score Hyperparameters" << std::endl;
output << "=== " << std::string(maxModel, '=') << " ========== ======== =========== " << std::string(maxHyper, '=') << std::endl;
numLines = 4;
auto i = 0;
for (const auto& item : data["results"]) {
auto color = (i % 2) ? Colors::BLUE() : Colors::CYAN();
auto score = item["score"].get<double>();
color = score == data["max_models"][item["model"].get<std::string>()] ? Colors::YELLOW() : color;
color = score == maxResult ? Colors::RED() : color;
output << color << std::setw(3) << std::fixed << std::right << i++ << " ";
output << std::setw(maxModel) << std::left << item["model"].get<std::string>() << " ";
output << color << item["date"].get<std::string>() << " ";
output << color << item["time"].get<std::string>() << " ";
output << std::setw(11) << std::setprecision(9) << std::fixed << score << " ";
output << item["hyperparameters"].get<std::string>() << std::endl;
numLines++;
}
data["dataset"] = dataset;
data["score"] = score;
data["model"] = model;
data["lengths"]["maxModel"] = maxModel;
data["lengths"]["maxHyper"] = maxHyper;
data["maxResult"] = maxResult;
}
}

27
src/results/ResultsDatasetConsole.h Normal file
View File

@@ -0,0 +1,27 @@
#pragma once
#include <locale>
#include <string>
#include <sstream>
#include <nlohmann/json.hpp>
#include "results/ResultsDataset.h"
namespace platform {
class ResultsDatasetsConsole {
public:
ResultsDatasetsConsole() = default;
~ResultsDatasetsConsole() = default;
std::string getOutput() const { return output.str(); }
int getNumLines() const { return numLines; }
json getData() { return data; }
void list_results(const std::string& dataset, const std::string& score, const std::string& model);
private:
std::stringstream output;
json data;
int numLines = 0;
};
}

95
src/results/ResultsDatasetConsole.hpp
View File

@@ -1,95 +0,0 @@
#pragma once
#include <locale>
#include <map>
#include <sstream>
#include <nlohmann/json.hpp>
#include "common/Colors.h"
#include "results/ResultsDataset.h"
namespace platform {
class ResultsDatasetsConsole {
public:
ResultsDatasetsConsole() = default;
~ResultsDatasetsConsole() = default;
std::string getOutput() const { return output.str(); }
int getNumLines() const { return numLines; }
json& getData() { return data; }
void list_results(const std::string& dataset, const std::string& score, const std::string& model)
{
auto results = platform::ResultsDataset(dataset, model, score);
results.load();
results.sortModel();
if (results.empty()) {
std::cerr << Colors::RED() << "No results found for dataset " << dataset << " and model " << model << Colors::RESET() << std::endl;
exit(1);
}
int maxModel = results.maxModelSize();
int maxHyper = results.maxHyperSize();
double maxResult = results.maxResultScore();
// Build data for the Report
json data = json::object();
data["results"] = json::array();
data["max_models"] = json::object(); // Max score per model
for (const auto& result : results) {
auto results = result.getData();
if (!data["max_models"].contains(result.getModel())) {
data["max_models"][result.getModel()] = 0;
}
for (const auto& item : results["results"]) {
if (item["dataset"] == dataset) {
// Store data for Excel report
json res = json::object();
res["date"] = result.getDate();
res["time"] = result.getTime();
res["model"] = result.getModel();
res["score"] = item["score"].get<double>();
res["hyperparameters"] = item["hyperparameters"].dump();
data["results"].push_back(res);
if (item["score"].get<double>() > data["max_models"][result.getModel()]) {
data["max_models"][result.getModel()] = item["score"].get<double>();
}
break;
}
}
}
//
// List the results
//
output << Colors::GREEN() << "Results of dataset " << dataset << " - for " << model << " model" << std::endl;
output << "There are " << results.size() << " results" << std::endl;
output << Colors::GREEN() << " # " << std::setw(maxModel + 1) << std::left << "Model" << "Date Time Score Hyperparameters" << std::endl;
output << "=== " << std::string(maxModel, '=') << " ========== ======== =========== " << std::string(maxHyper, '=') << std::endl;
numLines = 4;
auto i = 0;
for (const auto& item : data["results"]) {
auto color = (i % 2) ? Colors::BLUE() : Colors::CYAN();
auto score = item["score"].get<double>();
color = score == data["max_models"][item["model"].get<std::string>()] ? Colors::YELLOW() : color;
color = score == maxResult ? Colors::RED() : color;
output << color << std::setw(3) << std::fixed << std::right << i++ << " ";
output << std::setw(maxModel) << std::left << item["model"].get<std::string>() << " ";
output << color << item["date"].get<std::string>() << " ";
output << color << item["time"].get<std::string>() << " ";
output << std::setw(11) << std::setprecision(9) << std::fixed << score << " ";
output << item["hyperparameters"].get<std::string>() << std::endl;
numLines++;
}
data["dataset"] = dataset;
data["score"] = score;
data["model"] = model;
data["lengths"]["maxModel"] = maxModel;
data["lengths"]["maxHyper"] = maxHyper;
data["maxResult"] = maxResult;
}
private:
std::stringstream output;
json data;
int numLines = 0;
};
}