Class for holding an EventProcessor's histogram pointers and making sure that they all end up in the same directory in the output histogram file. More...

#include <HistogramPool.h>

Public Member Functions
	HistogramPool (std::function< TDirectory *()> get_directory)
	define how we can get the directory we need

void	setWeight (double w)
	Set the weight for filling the histograms.

TH1 *	get (const std::string &name)
	get a histogram from this pool by name

void	create (const config::Parameters &p)
	Create a histogram from the input configuration parameters.

void	create (const std::string &name, const std::string &x_label, const int &bins, const double &xmin, const double &xmax, bool weighted=false)
	Create a ROOT 1D histogram with uniform binning.

void	create (const std::string &name, const std::string &x_label, const std::vector< double > &bins, bool weighted=false)
	Create a ROOT 1D histogram with variable binning.

void	create (const std::string &name, const std::string &x_label, const std::vector< std::string > &categories, bool weighted=false)
	Create a ROOT 1D histogram with categorical binning.

void	create (const std::string &name, const std::string &x_label, const int &xbins, const double &xmin, const double &xmax, const std::string &y_label, const int &ybins, const double &ymin, const double &ymax, bool weighted=false)
	Create a ROOT 2D histogram with both axes having uniform binning.

void	create (const std::string &name, const std::string &x_label, const std::vector< double > &xbins, const std::string &y_label, const int &ybins, const double &ymin, const double &ymax, bool weighted=false)
	Create a ROOT 2D histogram with variable bins on the x axis and uniform bins on the y axis.

void	create (const std::string &name, const std::string &x_label, const std::vector< std::string > &xcategories, const std::string &y_label, const int &ybins, const double &ymin, const double &ymax, bool weighted=false)
	Create a ROOT 2D histogram with category bins on the x axis and uniform bins on the y axis.

void	create (const std::string &name, const std::string &x_label, const int &xbins, const double &xmin, const double &xmax, const std::string &y_label, const std::vector< double > &ybins, bool weighted=false)
	Create a ROOT 2D histogram with uniform bins on the x axis and variable bins on the y axis.

void	create (const std::string &name, const std::string &x_label, const std::vector< double > &xbins, const std::string &y_label, const std::vector< double > &ybins, bool weighted=false)
	Create a ROOT 2D histogram with variable bins on the x axis and variable bins on the y axis.

void	create (const std::string &name, const std::string &x_label, const std::vector< std::string > &xcategories, const std::string &y_label, const std::vector< double > &ybins, bool weighted=false)
	Create a ROOT 2D histogram with category bins on the x axis and variable bins on the y axis.

void	create (const std::string &name, const std::string &x_label, const int &xbins, const double &xmin, const double &xmax, const std::string &y_label, const std::vector< std::string > &ycategories, bool weighted=false)
	Create a ROOT 2D histogram with uniform bins on the x axis and category bins on the y axis.

void	create (const std::string &name, const std::string &x_label, const std::vector< double > &xbins, const std::string &y_label, const std::vector< std::string > &ycategories, bool weighted=false)
	Create a ROOT 2D histogram with variable bins on the x axis and category bins on the y axis.

void	create (const std::string &name, const std::string &x_label, const std::vector< std::string > &xcategories, const std::string &y_label, const std::vector< std::string > &ycategories, bool weighted=false)
	Create a ROOT 2D histogram with category bins on the x axis and category bins on the y axis.

template<typename T >
void	fill (const std::string &name, const T &val)
	Fill a 1D histogram.

template<typename T >
void	fillw (const std::string &name, const T &val, double w)
	Fill a 1D histogram.

template<typename Tx , typename Ty >
void	fill (const std::string &name, const Tx &valx, const Ty &valy)
	Fill a 2D histogram.

template<typename Tx , typename Ty >
void	fillw (const std::string &name, const Tx &valx, const Ty &valy, double w)
	Fill a 2D histogram.

Private Member Functions
void	insert (const std::string &name, std::function< TH1 *()> factory, bool weighted)
	insert a histogram into this pool by name

Private Attributes
double	the_weight_ {1.}
	The weight to fill histograms with.

std::unordered_map< std::string, TH1 * >	histograms_
	the pool of histogram pointers

std::function< TDirectory *()>	get_directory_
	the callback to get the directory these histograms should go in

Detailed Description

Class for holding an EventProcessor's histogram pointers and making sure that they all end up in the same directory in the output histogram file.

Note: In v4.5.1 of ldmx-sw and earlier, all of the processors shared the same pool of histograms, so the naming was a little funky. If a processor's name was "myProc" and a histogram's name was "myHist", the histogram would be written to "myProc/myProc_myHist" in the output histogram file. After v4.5.2, this central pool was abandoned and the output histograms are written to the more obvious location "myProc/myHist" in the output histogram file.

Usage

Each of the EventProcessors have a histograms_ member variable that developers can use within their producers and analyzers. You need to create the histograms before filling them. There are three different ways to specify the bins of a histogram:

uniform bins: provide the number of bins, the minimum, and the maximum
variable bins: provide the full list of bin edges (as a std::vector)
categories: provide a list of named categories (as a std::vector<std::string>)

With these three different ways to specify bins, there are three different ways to create a 1D histogram and nine different ways to create a 2D histogram. (Future developer note: if we want to increase the number of ways to specify the bins – "axis type" – * or increase the number of dimensions again, we probably just want to use a builder pattern like Boost.Histogram or just use Boost.Histogram under-the-hood and copy the resulting histograms into ROOT histograms at end of run time.)

In onProcessStart() or in the Python configuration, you create the histograms that you will want to fill. For example,

histograms_.create("my_variable", "Label for Axis", 10, 0.0, 1.0);

framework::HistogramPool::histograms_

std::unordered_map< std::string, TH1 * > histograms_

the pool of histogram pointers

Definition HistogramPool.h:112

or

analyzer.build1DHistogram("my_variable", "Label for Axis", 10, 0.0, 1.0)

And then in analyze or produce you fill the histograms.

histograms_.fill("my_variable", the_value);

Attempting to create histograms without the Python configuration providing an output histogram file will fail because then there is no place for the histograms to be saved.

The fill function uses the current setting of the weight in the histograms_object. This is helpful if, for example, you are putting one entry in the histogram for each event and you want the histogram to use the event weights. In order to do this event weighting for histograms, you would add the following line and all of the histograms that are filled after this line will use the event weight.

histograms_.setWeight(event.getEventWeight());

If you don't want to use special weighting, that's fine. The default weight is 1 and you may even have separate histograms in your processor that use different weights for example.

histograms_setWeight(1);
 
histograms_.fill("h_without_event_weight", value);
 
histograms_.setWeight(event.getEventWeight());
 
histograms_.fill("h_with_event_weight", value);

Additionally, if you want to use individual weights for the different histograms, you can use the fillw function to provide a weight when calling fill.

Note: By default, the histograms do not store the sum of the square weights in order to save space. If you are filling histograms with weights that are not 1 and you care about correct error estimates using those histograms, make sure to create the histogram with the weighted argument set to true.

Definition at line 107 of file HistogramPool.h.

Constructor & Destructor Documentation

◆ HistogramPool()

framework::HistogramPool::HistogramPool ( std::function< TDirectory *()> get_directory )

inline

define how we can get the directory we need

Definition at line 147 of file HistogramPool.h.

148 : get_directory_{get_directory} {}

framework::HistogramPool::get_directory_

std::function< TDirectory *()> get_directory_

the callback to get the directory these histograms should go in

Definition HistogramPool.h:123

Member Function Documentation

◆ create() [1/13]

void framework::HistogramPool::create ( const config::Parameters & p )

Create a histogram from the input configuration parameters.

Definition at line 69 of file HistogramPool.cxx.

                                                    {
  auto name{p.get<std::string>("name")};
  auto x_label{p.get<std::string>("xlabel")};
  auto y_label{p.get<std::string>("ylabel")};
  auto weighted{p.get<bool>("weighted")};
  auto numeric_xbins{p.get<std::vector<double>>("xbins")};
  auto category_xbins{p.get<std::vector<std::string>>("xcategories", {})};
  auto numeric_ybins{p.get<std::vector<double>>("ybins", {})};
  auto category_ybins{p.get<std::vector<std::string>>("ycategories", {})};
 
  bool one_dim = (numeric_ybins.empty() and category_ybins.empty());
  bool x_is_category = (not category_xbins.empty());
  bool y_is_category = (not category_ybins.empty());
  if (one_dim) {
    // assume 1D histogram
    if (x_is_category) {
      create(name, x_label, category_xbins, weighted);
    } else {
      create(name, x_label, numeric_xbins, weighted);
    }
  } else {
    if (x_is_category and y_is_category) {
      create(name, x_label, category_xbins, y_label, category_ybins, weighted);
    } else if (x_is_category and not y_is_category) {
      create(name, x_label, category_xbins, y_label, numeric_ybins, weighted);
    } else if (not x_is_category and y_is_category) {
      create(name, x_label, numeric_xbins, y_label, category_ybins, weighted);
    } else /* not x_is_category and not y_is_category */ {
      create(name, x_label, numeric_xbins, y_label, numeric_ybins, weighted);
    }
  }
}

References create(), and framework::config::Parameters::get().

Referenced by create(), dqm::SimObjects::createCalorimeterHists(), dqm::SimObjects::createTrackerHists(), dqm::HCalRawDigi::onProcessStart(), dqm::SimObjects::onProcessStart(), dqm::Trigger::onProcessStart(), dqm::TrigScintClusterDQM::onProcessStart(), dqm::TrigScintDQM::onProcessStart(), dqm::TrigScintHitDQM::onProcessStart(), dqm::TrigScintTrackDQM::onProcessStart(), ldmx::ecal::TrigPrimResolutionAnalyzer::onProcessStart(), Standalone::onProcessStart(), and TEST_CASE().

◆ create() [2/13]

void framework::HistogramPool::create	(	const std::string &	name,
		const std::string &	x_label,
		const int &	bins,
		const double &	xmin,
		const double &	xmax,
		bool	weighted = false )

Create a ROOT 1D histogram with uniform binning.

Parameters

name	Name of the histogram.
x_label	Title of the x axis.
bins	Total number of histogram bins.
xmin	The lower histogram limit.
xmax	The upper histogram limit.
weighted	whether to track the sum of squared weights separately

Definition at line 117 of file HistogramPool.cxx.

                                                              {
  insert(
      name,
      [&]() {
        auto hist = new TH1F(name.c_str(), "", bins, xmin, xmax);
        hist->GetXaxis()->SetTitle(x_label.c_str());
        return hist;
      },
      weighted);
}

References insert().

◆ create() [3/13]

void framework::HistogramPool::create	(	const std::string &	name,
		const std::string &	x_label,
		const int &	xbins,
		const double &	xmin,
		const double &	xmax,
		const std::string &	y_label,
		const int &	ybins,
		const double &	ymin,
		const double &	ymax,
		bool	weighted = false )

Create a ROOT 2D histogram with both axes having uniform binning.

Parameters

name	Name of the histogram.
x_label	Title of the x axis.
xbins	Total number of histogram bins in x.
xmin	The lower histogram limit in x.
xmax	The upper histogram limit in x.
y_label	Title of the x axis.
ybins	Total number of histogram bins in y.
ymin	The lower histogram limit in y.
ymax	The upper histogram limit in y.
weighted	whether to track the sum of squared weights separately

Definition at line 142 of file HistogramPool.cxx.

                                                              {
  insert(
      name,
      [&]() {
        auto hist =
            new TH2F(name.c_str(), "", xbins, xmin, xmax, ybins, ymin, ymax);
        hist->GetXaxis()->SetTitle(x_label.c_str());
        hist->GetYaxis()->SetTitle(y_label.c_str());
        return hist;
      },
      weighted);
}

References insert().

◆ create() [4/13]

void framework::HistogramPool::create	(	const std::string &	name,
		const std::string &	x_label,
		const int &	xbins,
		const double &	xmin,
		const double &	xmax,
		const std::string &	y_label,
		const std::vector< double > &	ybins,
		bool	weighted = false )

Create a ROOT 2D histogram with uniform bins on the x axis and variable bins on the y axis.

Parameters

name	Name of the histogram.
x_label	Title of the x axis.
xbins	Total number of histogram bins in x.
xmin	The lower histogram limit in x.
xmax	The upper histogram limit in x.
y_label	Title of the x axis.
ybins	vector of bin edges
weighted	whether to track the sum of squared weights separately

Definition at line 195 of file HistogramPool.cxx.

                                                                          {
  insert(
      name,
      [&]() {
        auto hist = new TH2F(name.c_str(), "", xbins, xmin, xmax,
                             ybins.size() - 1, ybins.data());
        hist->GetXaxis()->SetTitle(x_label.c_str());
        hist->GetYaxis()->SetTitle(y_label.c_str());
        return hist;
      },
      weighted);
}

References insert().

◆ create() [5/13]

void framework::HistogramPool::create	(	const std::string &	name,
		const std::string &	x_label,
		const int &	xbins,
		const double &	xmin,
		const double &	xmax,
		const std::string &	y_label,
		const std::vector< std::string > &	ycategories,
		bool	weighted = false )

Create a ROOT 2D histogram with uniform bins on the x axis and category bins on the y axis.

Parameters

name	Name of the histogram.
x_label	Title of the x axis.
xbins	Total number of histogram bins in x.
xmin	The lower histogram limit in x.
xmax	The upper histogram limit in x.
y_label	Title of the x axis.
ycategories	vector of string categories
weighted	whether to track the sum of squared weights separately

Definition at line 211 of file HistogramPool.cxx.

                                          {
  insert(
      name,
      [&]() {
        auto [nybins, ymin, ymax] = categoryBins(ycategories);
        auto hist =
            new TH2F(name.c_str(), "", xbins, xmin, xmax, nybins, ymin, ymax);
        labelAxis(hist->GetYaxis(), ycategories);
        hist->GetXaxis()->SetTitle(x_label.c_str());
        hist->GetYaxis()->SetTitle(y_label.c_str());
        return hist;
      },
      weighted);
}

References insert().

◆ create() [6/13]

void framework::HistogramPool::create	(	const std::string &	name,
		const std::string &	x_label,
		const std::vector< double > &	bins,
		bool	weighted = false )

Create a ROOT 1D histogram with variable binning.

Parameters

name	Name of the histogram.
x_label	Title of the x axis.
bins	vector of bin edges
weighted	whether to track the sum of squared weights separately

Definition at line 130 of file HistogramPool.cxx.

                                                                         {
  insert(
      name,
      [&]() {
        auto hist = new TH1F(name.c_str(), "", bins.size() - 1, bins.data());
        hist->GetXaxis()->SetTitle(x_label.c_str());
        return hist;
      },
      weighted);
}

References insert().

◆ create() [7/13]

void framework::HistogramPool::create	(	const std::string &	name,
		const std::string &	x_label,
		const std::vector< double > &	xbins,
		const std::string &	y_label,
		const int &	ybins,
		const double &	ymin,
		const double &	ymax,
		bool	weighted = false )

Create a ROOT 2D histogram with variable bins on the x axis and uniform bins on the y axis.

Parameters

name	Name of the histogram.
x_label	Title of the x axis.
xbins	vector of bin edges
y_label	Title of the x axis.
ybins	Total number of histogram bins in y.
ymin	The lower histogram limit in y.
ymax	The upper histogram limit in y.
weighted	whether to track the sum of squared weights separately

Definition at line 159 of file HistogramPool.cxx.

                                          {
  insert(
      name,
      [&]() {
        auto hist = new TH2F(name.c_str(), "", xbins.size() - 1, xbins.data(),
                             ybins, ymin, ymax);
        hist->GetXaxis()->SetTitle(x_label.c_str());
        hist->GetYaxis()->SetTitle(y_label.c_str());
        return hist;
      },
      weighted);
}

References insert().

◆ create() [8/13]

void framework::HistogramPool::create	(	const std::string &	name,
		const std::string &	x_label,
		const std::vector< double > &	xbins,
		const std::string &	y_label,
		const std::vector< double > &	ybins,
		bool	weighted = false )

Create a ROOT 2D histogram with variable bins on the x axis and variable bins on the y axis.

Parameters

name	Name of the histogram.
x_label	Title of the x axis.
xbins	vector of bin edges
y_label	Title of the x axis.
ybins	vector of bin edges
weighted	whether to track the sum of squared weights separately

Definition at line 230 of file HistogramPool.cxx.

                                                                          {
  insert(
      name,
      [&]() {
        auto hist = new TH2F(name.c_str(), "", xbins.size() - 1, xbins.data(),
                             ybins.size() - 1, ybins.data());
        hist->GetXaxis()->SetTitle(x_label.c_str());
        hist->GetYaxis()->SetTitle(y_label.c_str());
        return hist;
      },
      weighted);
}

References insert().

◆ create() [9/13]

void framework::HistogramPool::create	(	const std::string &	name,
		const std::string &	x_label,
		const std::vector< double > &	xbins,
		const std::string &	y_label,
		const std::vector< std::string > &	ycategories,
		bool	weighted = false )

Create a ROOT 2D histogram with variable bins on the x axis and category bins on the y axis.

Parameters

name	Name of the histogram.
x_label	Title of the x axis.
xbins	vector of bin edges
y_label	Title of the x axis.
ycategories	vector of string categories
weighted	whether to track the sum of squared weights separately

Definition at line 246 of file HistogramPool.cxx.

                                          {
  insert(
      name,
      [&]() {
        auto [nybins, ymin, ymax] = categoryBins(ycategories);
        auto hist = new TH2F(name.c_str(), "", xbins.size() - 1, xbins.data(),
                             nybins, ymin, ymax);
        labelAxis(hist->GetYaxis(), ycategories);
        hist->GetXaxis()->SetTitle(x_label.c_str());
        hist->GetYaxis()->SetTitle(y_label.c_str());
        return hist;
      },
      weighted);
}

References insert().

◆ create() [10/13]

void framework::HistogramPool::create	(	const std::string &	name,
		const std::string &	x_label,
		const std::vector< std::string > &	categories,
		bool	weighted = false )

Create a ROOT 1D histogram with categorical binning.

Parameters

name	Name of the histogram.
x_label	Title of the x axis.
categories	list of string categories to name the bins
weighted	whether to track the sum of squared weights separately

Definition at line 102 of file HistogramPool.cxx.

                                          {
  insert(
      name,
      [&]() {
        auto [nbins, xmin, xmax] = categoryBins(categories);
        auto hist = new TH1F(name.c_str(), "", nbins, xmin, xmax);
        hist->GetXaxis()->SetTitle(x_label.c_str());
        labelAxis(hist->GetXaxis(), categories);
        return hist;
      },
      weighted);
}

References insert().

◆ create() [11/13]

void framework::HistogramPool::create	(	const std::string &	name,
		const std::string &	x_label,
		const std::vector< std::string > &	xcategories,
		const std::string &	y_label,
		const int &	ybins,
		const double &	ymin,
		const double &	ymax,
		bool	weighted = false )

Create a ROOT 2D histogram with category bins on the x axis and uniform bins on the y axis.

Parameters

name	Name of the histogram.
x_label	Title of the x axis.
xcategories	vector of string categories
y_label	Title of the x axis.
ybins	Total number of histogram bins in y.
ymin	The lower histogram limit in y.
ymax	The upper histogram limit in y.
weighted	whether to track the sum of squared weights separately

Definition at line 176 of file HistogramPool.cxx.

                                          {
  insert(
      name,
      [&]() {
        auto [nxbins, xmin, xmax] = categoryBins(xcategories);
        auto hist =
            new TH2F(name.c_str(), "", nxbins, xmin, xmax, ybins, ymin, ymax);
        labelAxis(hist->GetXaxis(), xcategories);
        hist->GetXaxis()->SetTitle(x_label.c_str());
        hist->GetYaxis()->SetTitle(y_label.c_str());
        return hist;
      },
      weighted);
}

References insert().

◆ create() [12/13]

void framework::HistogramPool::create	(	const std::string &	name,
		const std::string &	x_label,
		const std::vector< std::string > &	xcategories,
		const std::string &	y_label,
		const std::vector< double > &	ybins,
		bool	weighted = false )

Create a ROOT 2D histogram with category bins on the x axis and variable bins on the y axis.

Parameters

name	Name of the histogram.
x_label	Title of the x axis.
xcategories	vector of string categories
y_label	Title of the x axis.
ybins	vector of bin edges
weighted	whether to track the sum of squared weights separately

Definition at line 265 of file HistogramPool.cxx.

                                                                          {
  insert(
      name,
      [&]() {
        auto [nxbins, xmin, xmax] = categoryBins(xcategories);
        auto hist = new TH2F(name.c_str(), "", nxbins, xmin, xmax,
                             ybins.size() - 1, ybins.data());
        labelAxis(hist->GetXaxis(), xcategories);
        hist->GetXaxis()->SetTitle(x_label.c_str());
        hist->GetYaxis()->SetTitle(y_label.c_str());
        return hist;
      },
      weighted);
}

References insert().

◆ create() [13/13]

void framework::HistogramPool::create	(	const std::string &	name,
		const std::string &	x_label,
		const std::vector< std::string > &	xcategories,
		const std::string &	y_label,
		const std::vector< std::string > &	ycategories,
		bool	weighted = false )

Create a ROOT 2D histogram with category bins on the x axis and category bins on the y axis.

Parameters

name	Name of the histogram.
x_label	Title of the x axis.
xcategories	vector of string categories
y_label	Title of the x axis.
ycategories	vector of string categories
weighted	whether to track the sum of squared weights separately

Definition at line 283 of file HistogramPool.cxx.

                                          {
  insert(
      name,
      [&]() {
        auto [nybins, ymin, ymax] = categoryBins(ycategories);
        auto [nxbins, xmin, xmax] = categoryBins(xcategories);
        auto hist =
            new TH2F(name.c_str(), "", nxbins, xmin, xmax, nybins, ymin, ymax);
        labelAxis(hist->GetYaxis(), ycategories);
        labelAxis(hist->GetXaxis(), xcategories);
        hist->GetXaxis()->SetTitle(x_label.c_str());
        hist->GetYaxis()->SetTitle(y_label.c_str());
        return hist;
      },
      weighted);
}

References insert().

◆ fill() [1/2]

template<typename T >

void framework::HistogramPool::fill	(	const std::string &	name,
		const T &	val )

inline

Fill a 1D histogram.

Uses the current setting of the weight.

Parameters

name	name of the histogram to fill
val	value to fill

Definition at line 367 of file HistogramPool.h.

                                                 {
    auto hist = dynamic_cast<TH1F*>(this->get(name));
    if (hist) {
      hist->Fill(val, the_weight_);
    } else {
      // the `get` method handles checking if the histogram exists
      // the only way that hist would be null at this point is if its not a TH1F
      EXCEPTION_RAISE(
          "BadHistSize",
          "Attempting to 1D fill a histogram that is not actually 1D.");
    }
  }

References get(), and the_weight_.

◆ fill() [2/2]

template<typename Tx , typename Ty >

void framework::HistogramPool::fill	(	const std::string &	name,
		const Tx &	valx,
		const Ty &	valy )

inline

Fill a 2D histogram.

Uses the current setting of the weight.

Parameters

name	name of the histogram to fill
valx	x value to fill
valy	y value to fill

Definition at line 413 of file HistogramPool.h.

                                                                   {
    auto hist = dynamic_cast<TH2F*>(this->get(name));
    if (hist) {
      hist->Fill(valx, valy, the_weight_);
    } else {
      // the `get` method handles checking if the histogram exists
      // the only way that hist would be null at this point is if its not a TH2F
      EXCEPTION_RAISE(
          "BadHistSize",
          "Attempting to 2D fill a histogram that is not actually 2D.");
    }
  }

References get(), and the_weight_.

◆ fillw() [1/2]

template<typename T >

void framework::HistogramPool::fillw	(	const std::string &	name,
		const T &	val,
		double	w )

inline

Fill a 1D histogram.

Using the input weight.

Parameters

name	name of the histogram to fill
val	value to fill
w	weight to fill with

Definition at line 390 of file HistogramPool.h.

                                                            {
    auto hist = dynamic_cast<TH1F*>(this->get(name));
    if (hist) {
      hist->Fill(val, w);
    } else {
      // the `get` method handles checking if the histogram exists
      // the only way that hist would be null at this point is if its not a TH1F
      EXCEPTION_RAISE(
          "BadHistSize",
          "Attempting to 1D fill a histogram that is not actually 1D.");
    }
  }

References get().

◆ fillw() [2/2]

template<typename Tx , typename Ty >

void framework::HistogramPool::fillw	(	const std::string &	name,
		const Tx &	valx,
		const Ty &	valy,
		double	w )

inline

Fill a 2D histogram.

Using the input weight.

Parameters

name	name of the histogram to fill
valx	x value to fill
valy	y value to fill
w	weight to fill with

Definition at line 437 of file HistogramPool.h.

                       {
    auto hist = dynamic_cast<TH2F*>(this->get(name));
    if (hist) {
      hist->Fill(valx, valy, w);
    } else {
      // the `get` method handles checking if the histogram exists
      // the only way that hist would be null at this point is if its not a TH2F
      EXCEPTION_RAISE(
          "BadHistSize",
          "Attempting to 2D fill a histogram that is not actually 2D.");
    }
  }

References get().

◆ get()

TH1 * framework::HistogramPool::get ( const std::string & name )

get a histogram from this pool by name

Definition at line 15 of file HistogramPool.cxx.

                                             {
  auto histo = histograms_.find(name);
  if (histo == histograms_.end()) {
    EXCEPTION_RAISE(
        "InvalidArg",
        "Histogram " + name +
            " not found in pool."
            "\nMake sure to `histograms_.create` in onProcessStart for any "
            "histogram you want to `histograms_.fill`.");
  }
 
  return histograms_[name];
}

References histograms_.

Referenced by tracking::dqm::RawSiStripDQM::analyze(), fill(), fill(), fillw(), fillw(), and tracking::dqm::RawSiStripDQM::getHists().

◆ insert()

void framework::HistogramPool::insert	(	const std::string &	name,
		std::function< TH1 *()>	factory,
		bool	weighted )

private

insert a histogram into this pool by name

Note: The fill method assumes we are creating TH1F or TH2F histograms which is what is implemented in the create methods.

Parameters

[in]	name	name to store histogram under in the pool, the `create` methods below ensure that the name in the pool and the name in the output file are the same
[in]	factory	the function that creates the histogram which is called after we go into the appropriate output directory, the returned pointer is then put into the pool after checking that it doesn't exist yet.
[in]	weighted	true if histogram is weighted (we then call Sumw2) or false otherwise

Exceptions

framework::Exception if the passed name already exists

The histogram needs to be created after we cd into the appropriate output directory which is why we pass in the factory function that creates the histogram into this function rather than a pointer to the already created histogram itself.

Definition at line 29 of file HistogramPool.cxx.

                                                                       {
  if (histograms_.find(name) != histograms_.end()) {
    EXCEPTION_RAISE(
        "RepeatName",
        "Histogram " + name +
            " already exists in histogram pool."
            "\nMake sure to use distinct names for your different histograms!"
            "\nYou can use `histograms_.get` to retrieve a pointer to a "
            "specific histogram "
            "if you want to do some other customizations besides filling.");
  }
 
  get_directory_()->cd();
 
  auto hist = factory();
  if (weighted) hist->Sumw2();
  histograms_[name] = hist;
}

References get_directory_, and histograms_.

Referenced by create(), create(), create(), create(), create(), create(), create(), create(), create(), create(), create(), and create().

◆ setWeight()

void framework::HistogramPool::setWeight ( double w )

inline

Set the weight for filling the histograms.

Definition at line 153 of file HistogramPool.h.

153{ the_weight_ = w; }

References the_weight_.

Referenced by dqm::DarkBremInteraction::produce(), and dqm::EcalSPElectronKinematics::produce().

Member Data Documentation

◆ get_directory_

std::function<TDirectory*()> framework::HistogramPool::get_directory_

private

the callback to get the directory these histograms should go in

This needs to be dynamic so that the directory is only created upon request.

Note: the returned TDirectory pointer is immediately de-referenced, so we will get a segmentation fault if the function stored here returns an invalid address (such as nullptr).

Definition at line 123 of file HistogramPool.h.

Referenced by insert().

◆ histograms_

std::unordered_map<std::string, TH1*> framework::HistogramPool::histograms_

private

the pool of histogram pointers

Definition at line 112 of file HistogramPool.h.

Referenced by get(), and insert().

◆ the_weight_

double framework::HistogramPool::the_weight_ {1.}

private

The weight to fill histograms with.

Definition at line 110 of file HistogramPool.h.

110{1.};

Referenced by fill(), fill(), and setWeight().

The documentation for this class was generated from the following files:

Framework/include/Framework/HistogramPool.h
Framework/src/Framework/HistogramPool.cxx

Public Member Functions

Private Member Functions

Private Attributes

Detailed Description

Usage

Constructor & Destructor Documentation

◆ HistogramPool()

Member Function Documentation

◆ create() [1/13]

◆ create() [2/13]

◆ create() [3/13]

◆ create() [4/13]

◆ create() [5/13]

◆ create() [6/13]

◆ create() [7/13]

◆ create() [8/13]

◆ create() [9/13]

◆ create() [10/13]

◆ create() [11/13]

◆ create() [12/13]

◆ create() [13/13]

◆ fill() [1/2]

◆ fill() [2/2]

◆ fillw() [1/2]

◆ fillw() [2/2]

◆ get()

◆ insert()

◆ setWeight()

Member Data Documentation

◆ get_directory_

◆ histograms_

◆ the_weight_