openstudio-standards 0.8.5 → 0.8.7

data/lib/openstudio-standards/btap/qaqc/datapoint.rb

@@ -0,0 +1,641 @@
+require 'openstudio'
+require 'securerandom'
+require 'yaml'
+
+module BTAP
+  class Datapoint
+    def initialize(input_folder:   File.join(__dir__, 'input'),
+                   output_folder:  File.join(__dir__, 'output'),
+                   weather_folder: File.join(__dir__, 'weather'))
+
+      @failed = false
+
+      puts("INPUT FOLDER:  #{input_folder}\n" \
+           "OUTPUT FOLDER: #{output_folder}")
+
+      # Set local temporary cache folders. These are created new for each
+      # datapoint. Location of the datapoint input folder. Could be local, cloud,
+      # or VM host.
+      datapoint_folder = File.join(__dir__, 'datapoint/')
+      @dp_temp_folder  = File.join(datapoint_folder, SecureRandom.uuid + '_' + Process.pid.to_s)
+
+      # Determine whether the analysis was run on AWS S3.
+      @s3 = input_folder.start_with?('s3:')
+
+      # Clean up stale temp folders whose PID isn't running.
+      Dir.glob("#{datapoint_folder}/*").each do |dir|
+        begin
+          Process.kill(0, dir.rpartition('_').last.to_i)
+        rescue Errno::ESRCH, Errno::EPERM
+          FileUtils.rm_rf(dir)
+        end
+      end
+      FileUtils.mkdir_p(@dp_temp_folder)
+
+      # For S3 inputs, download to the temp folder so files are accessible
+      # locally. For local inputs, read directly from the input folder.
+      if @s3
+        match = input_folder.match(%r{s3://(.*?)/(.*)})
+        @s3_bucket = match[1]
+        @s3_input_folder_object = match[2]
+        @s3_output_folder_object = output_folder.match(%r{s3://(.*?)/(.*)})[2]
+        puts("s3_bucket: #{@s3_bucket}\n" \
+             "s3_input_folder_object: #{@s3_input_folder_object}\n" \
+             "s3_output_folder_object: #{@s3_output_folder_object}")
+
+        s3_copy(source_folder: input_folder, target_folder: @dp_temp_folder)
+        run_options_path = File.join(@dp_temp_folder, 'run_options.yml')
+      else
+        raise("Input folder does not exist: #{input_folder}") unless Dir.exist?(input_folder)
+        run_options_path = File.join(input_folder, 'run_options.yml')
+      end
+      raise("Could not read input from #{run_options_path}") unless File.file?(run_options_path)
+
+      @options = YAML.load_file(run_options_path)
+
+      # Location of datapoint output folder. Could be local, cloud, or VM host.
+      @dp_output_folder = File.join(output_folder, @options[:datapoint_id])
+
+      # Show versions in YAML file.
+      @options[:os_git_revision] = OpenstudioStandards.git_revision
+
+      # Get users' inputs for the parameters needed for the calculation af net present value
+      @npv_start_year = @options[:npv_start_year]
+      @npv_end_year = @options[:npv_end_year]
+      @npv_discount_rate = @options[:npv_discount_rate]
+      @npv_discount_rate_carbon = @options[:npv_discount_rate_carbon]
+
+      # Save configuration to temp folder.
+      File.open(File.join(@dp_temp_folder, 'run_options.yml'), 'w') { |file| file.write(@options.to_yaml) }
+      begin
+        @options[:template] = 'NECB2011' if @options[:algorithm_type] == 'osm_batch'
+        @standard = Standard.build(@options[:template])
+
+        # Set use the convert_arg_to_bool method from the standard class to set the @oerd_utility_pricing_flag
+        # Always set oerd_utility_pricing to false until it can be shared freely.
+        @oerd_utility_pricing = false
+        @utility_pricing_year = @standard.convert_arg_to_f(variable: @options[:utility_pricing_year], default: 2020)
+
+        # This allows you to select the skeleton model from our built-in starting
+        # points. You can add a custom file as it will search the libary first.
+        model = @standard.load_building_type_from_library(building_type: @options[:building_type])
+        if false == model
+          osm_model_path = File.absolute_path(File.join(@s3 ? @dp_temp_folder : input_folder, @options[:building_type] + '.osm'))
+          raise("File #{osm_model_path} not found") unless File.exist?(osm_model_path)
+
+          model = BTAP::FileIO.load_osm(osm_model_path)
+        end
+
+        # If analysis type is `osm_batch`, do nothing to the OSM files other than
+        # add temporal outputs and weather data. Otherwise, modify OSM input with
+        # options.
+        if @options[:algorithm_type] == 'osm_batch'
+          @standard.set_output_variables(model: model, output_variables: @options[:output_variables])
+          @standard.set_output_meters(model: model, output_meters: @options[:output_meters])
+          model.getYearDescription.setDayofWeekforStartDay('Sunday')
+          OpenstudioStandards::Weather.model_set_building_location(
+            model,
+            weather_file_path: OpenstudioStandards::Weather.get_standards_weather_file_path(@options[:epw_file]))
+        else
+          @standard.model_apply_standard(
+            model: model,
+            construction_opt: @options[:construction_opt],
+            epw_file: @options[:epw_file],
+            custom_weather_folder: weather_folder,
+            btap_weather: @options[:btap_weather],
+            sizing_run_dir: File.join(@dp_temp_folder, 'sizing_folder'),
+            hvac_system_primary: @options[:hvac_system_primary],
+            hvac_system_dwelling_units: @options[:hvac_system_dwelling_units],
+            hvac_system_washrooms: @options[:hvac_system_washrooms],
+            hvac_system_corridor: @options[:hvac_system_corridor],
+            hvac_system_storage: @options[:hvac_system_storage],
+            primary_heating_fuel: @options[:primary_heating_fuel],
+            necb_reference_hp: @options[:necb_reference_hp],
+            necb_reference_hp_supp_fuel: @options[:necb_reference_hp_supp_fuel],
+
+            # Four options: @options[: (1) 'NECB_Default', (2) 'No DCV',
+            # (3) 'Occupancy-based DCV' , (4) 'CO2-based DCV'
+            dcv_type: @options[:dcv_type],
+
+            # Two options: @options[: (1) 'NECB_Default', (2) 'LED'
+            lights_type: @options[:lights_type],
+            lights_scale: @options[:lights_scale],
+            daylighting_type: @options[:daylighting_type],
+
+            # Two options: @options[: (1) 'NECB_Default', (2) 'add_daylighting_controls'
+            ecm_system_name: @options[:ecm_system_name],
+            ecm_system_zones_map_option: @options[:ecm_system_zones_map_option],
+
+            # (1) 'NECB_Default' (2) 'one_sys_per_floor' (3) 'one_sys_per_bldg'
+            erv_package: @options[:erv_package],
+            boiler_eff: @options[:boiler_eff],
+
+            # TODO: Inconsistent naming
+            unitary_cop: @options[:adv_dx_units],
+            furnace_eff: @options[:furnace_eff],
+            shw_eff: @options[:shw_eff],
+            ext_wall_cond: @options[:ext_wall_cond],
+            ext_floor_cond: @options[:ext_floor_cond],
+            ext_roof_cond: @options[:ext_roof_cond],
+            ground_wall_cond: @options[:ground_wall_cond],
+            ground_floor_cond: @options[:ground_floor_cond],
+            ground_roof_cond: @options[:ground_roof_cond],
+            door_construction_cond: @options[:door_construction_cond],
+            fixed_window_cond: @options[:fixed_window_cond],
+            glass_door_cond: @options[:glass_door_cond],
+            overhead_door_cond: @options[:overhead_door_cond],
+            skylight_cond: @options[:skylight_cond],
+            glass_door_solar_trans: @options[:glass_door_solar_trans],
+            fixed_wind_solar_trans: @options[:fixed_wind_solar_trans],
+            skylight_solar_trans: @options[:skylight_solar_trans],
+            fdwr_set: @options[:fdwr_set],
+            srr_opt: @options[:srr_opt],
+            srr_set: @options[:srr_set],
+            rotation_degrees: @options[:rotation_degrees],
+            scale_x: @options[:scale_x],
+            scale_y: @options[:scale_y],
+            scale_z: @options[:scale_z],
+            footprint_aspect_ratio: @options[:footprint_aspect_ratio],
+            pv_ground_type: @options[:pv_ground_type],
+            pv_ground_total_area_pv_panels_m2: @options[:pv_ground_total_area_pv_panels_m2],
+            pv_ground_tilt_angle: @options[:pv_ground_tilt_angle],
+            pv_ground_azimuth_angle: @options[:pv_ground_azimuth_angle],
+            pv_ground_module_description: @options[:pv_ground_module_description],
+            nv_type: @options[:nv_type],
+            nv_opening_fraction: @options[:nv_opening_fraction],
+            nv_temp_out_min: @options[:nv_temp_out_min],
+            nv_delta_temp_in_out: @options[:nv_delta_temp_in_out],
+            occupancy_loads_scale: @options[:occupancy_loads_scale],
+            electrical_loads_scale: @options[:electrical_loads_scale],
+            oa_scale: @options[:oa_scale],
+            infiltration_scale: @options[:infiltration_scale],
+            chiller_type: @options[:chiller_type],
+            output_variables: @options[:output_variables],
+            output_meters: @options[:output_meters],
+            airloop_economizer_type: @options[:airloop_economizer_type],
+            shw_scale: @options[:shw_scale],
+            baseline_system_zones_map_option: @options[:baseline_system_zones_map_option],
+            tbd_interpolate: @options[:tbd_interpolate],
+            tbd_option: @options[:tbd_option],
+            necb_hdd: @options[:necb_hdd],
+            boiler_fuel: @options[:boiler_fuel],
+            boiler_cap_ratio: @options[:boiler_cap_ratio],
+            swh_fuel: @options[:swh_fuel],
+            airloop_fancoils_heating: @options[:airloop_fancoils_heating],
+            oerd_utility_pricing: @oerd_utility_pricing)
+        end
+
+        # Save model to disk.
+        puts "saving model to #{File.join(@dp_temp_folder, 'output.osm')}"
+        BTAP::FileIO.save_osm(model, File.join(@dp_temp_folder, 'output.osm'))
+
+        # Run annual simulation of model.
+        if @options[:run_annual_simulation]
+          @run_dir = File.join(@dp_temp_folder, 'run_dir')
+          puts "running simulation in #{@run_dir}"
+          @standard.model_run_simulation_and_log_errors(model, @run_dir)
+          raise("No SQL file found.") if model.sqlFile.empty?
+          # Create qaqc file and save it.
+          @qaqc = BTAP::Datapoint.build_qaqc(model, @standard, @options[:datapoint_id], @options[:analysis_id])
+
+          # Load the SQL file into model.
+          sql_path = OpenStudio::Path.new(File.join(@run_dir, 'run/eplusout.sql'))
+          unless OpenStudio.exists(sql_path)
+            raise(
+              OpenStudio::Error,
+              "openstudio.model.Model", "Results for the sizing run couldn't be found here: #{sql_path}.")
+          end
+
+          sql = OpenStudio::SqlFile.new(sql_path)
+
+          # Check to make sure the sql file is readable,
+          # which won't be true if EnergyPlus crashed during simulation.
+          unless sql.connectionOpen
+            raise(OpenStudio::Error, 'openstudio.model.Model', "The run failed. Look at the eplusout.err file in" \
+                  "#{File.dirname(sql_path.to_s)} to see the cause.")
+            return false
+          end
+
+          # Attach the SQL file from the run to the sizing model
+          model.setSqlFile(sql)
+
+          post_analysis = nil
+          if @options[:enable_costing] or @options[:enable_carbon]
+            post_analysis = BTAP::DatapointAnalysis.new(
+              model: model,
+              output_folder: @dp_temp_folder,
+              standard: @standard,
+              qaqc: @qaqc)
+          end
+
+          @cost_result = nil
+          if @options[:enable_costing]
+            @cost_result = post_analysis.run_costing(
+              costs_csv: Paths.costs_path,
+              factors_csv: Paths.costs_local_factors_path)
+          end
+
+          @carbon_result = nil
+          if @options[:enable_carbon]
+            @carbon_result = post_analysis.run_carbon
+          end
+
+          @qaqc[:options] = @options # Command-line options
+          @btap_data = BTAP::Data.new(model: model,
+                                    runner: nil,
+                                    cost_result: @cost_result,
+                                    carbon_result: @carbon_result,
+                                    qaqc: @qaqc,
+                                    npv_start_year: @npv_start_year,
+                                    npv_end_year: @npv_end_year,
+                                    npv_discount_rate: @npv_discount_rate,
+                                    npv_discount_rate_carbon: @npv_discount_rate_carbon,
+                                    oerd_utility_pricing: @oerd_utility_pricing,
+                                    utility_pricing_year: @utility_pricing_year).btap_data
+
+          File.open(File.join(@dp_temp_folder, 'btap_data.json'), 'w') do |file|
+            file.write(JSON.pretty_generate(@btap_data.sort.to_h, allow_nan: true))
+            puts "Wrote File btap_data.json in #{Dir.pwd} "
+          end
+
+          File.open(File.join(@dp_temp_folder, 'qaqc.json'), 'w') do |file|
+            file.write(JSON.pretty_generate(@qaqc, allow_nan: true))
+            puts "Wrote File qaqc.json in #{Dir.pwd} "
+          end
+
+          self.output_hourly_data(model,@dp_temp_folder, @options[:datapoint_id])
+          self.output_timestep_data(model,@dp_temp_folder, @options[:datapoint_id])
+        end
+      rescue StandardError => error
+        @error = error
+        puts "Error occured: #{@error}"
+        File.open(File.join(@dp_temp_folder, 'error.txt'), 'w') do |f|
+          f.write(@error.message + "\n" + @error.backtrace.join("\n"))
+        end
+        @failed = true
+      ensure
+        if @s3
+          @dp_output_folder = File.join(@s3_output_folder_object, @options[:datapoint_id])
+          s3_copy_folder_to_s3(bucket_name: @s3_bucket,
+                               source_folder: @dp_temp_folder,
+                               target_folder: @dp_output_folder)
+        else
+
+          # Copy results to datapoint output folder.
+          @dp_output_folder = File.join(output_folder, @options[:datapoint_id])
+          FileUtils.rm_rf(@dp_output_folder)
+          FileUtils.mkdir_p(@dp_output_folder)
+          FileUtils.cp_r(File.join(@dp_temp_folder, '.'), @dp_output_folder)
+          puts "Copied output to your designated output folder in the #{@options[:datapoint_id]} subfolder."
+        end
+
+        # Clean temp/cache folder up.
+        FileUtils.rm_rf(@dp_temp_folder)
+
+        # Do not fail container if running on AWS, handle error on AWS instead to
+        # avoid isuses with large analyses.
+        unless @failed == false || @s3
+          raise(@error)
+        end
+      end
+    end
+
+    class << self
+
+      # Initializes the qaqc data structure. Scoped inside of the class so
+      # that it can be used in the intialization of this class as well as in
+      # BTAP::Analysis.
+      def build_qaqc(model, standard, datapoint_id, analysis_id)
+        qaqc = standard.init_qaqc(model)
+        command = "SELECT Value
+                   FROM TabularDataWithStrings
+                   WHERE ReportName='LEEDsummary'
+                   AND ReportForString='Entire Facility'
+                   AND TableName='Sec1.1A-General Information'
+                   AND RowName = 'Principal Heating Source'
+                   AND ColumnName='Data'"
+        value = model.sqlFile.get.execAndReturnFirstString(command)
+        qaqc[:building][:principal_heating_source] = 'unknown'
+        unless value.empty?
+          qaqc[:building][:principal_heating_source] = value.get
+        end
+
+        if qaqc[:building][:principal_heating_source] == 'Additional Fuel'
+          model.getPlantLoops.sort.each do |iplantloop|
+            boilers = iplantloop.components.select { |icomponent| icomponent.to_BoilerHotWater.is_initialized }
+            qaqc[:building][:principal_heating_source] = 'FuelOilNo2' unless boilers.select { |boiler| boiler.to_BoilerHotWater.get.fuelType.to_s == 'FuelOilNo2' }.empty?
+          end
+        end
+
+        qaqc[:aws_datapoint_id] = datapoint_id
+        qaqc[:aws_analysis_id]  = analysis_id
+
+        return qaqc
+      end
+    end
+
+    def s3_copy_file_to_s3(bucket_name:, source_file:, target_file:, n: 0)
+      require 'aws-sdk-s3'
+      Aws.use_bundled_cert!
+      s3_client = Aws::S3::Client.new(region: 'ca-central-1')
+      transfer_manager = Aws::S3::TransferManager.new(client: s3_client)
+
+      puts("Copying File to S3. source_file:#{source_file} bucket:#{bucket_name} target_folder:#{target_file}")
+      response = nil
+      begin
+        result = transfer_manager.upload_file(source_file, bucket: bucket_name, key: target_file)
+
+        if result == true
+          puts "Object '#{source_file}' uploaded to bucket '#{bucket_name}'."
+        else
+          puts response
+          raise("Error:Object '#{source_file}' not uploaded to bucket '#{bucket_name}'.")
+        end
+      rescue StandardError => error
+
+        # Implementing exponential backoff timing out after 8 tries.
+        if n == 8
+          raise("Giving Up. Failed to submit send file #{source_file} #{target_file} in 8 tries while using exponential backoff. Error was:#{error}.")
+        end
+
+        wait_time = 2 ** n + rand
+        puts "Implementing exponential backoff for sending #{target_file} for #{wait_time} seconds. Attempt #{n}..."
+        sleep(wait_time)
+        return s3_copy_file_to_s3(bucket_name: bucket_name, source_file: source_file, target_file: target_file, n: n + 1)
+      end
+      return response
+    end
+
+    def s3_copy_folder_to_s3(bucket_name:, source_folder:, target_folder:)
+      puts("Copying Folder to S3. source_folder:#{source_folder} bucket:#{bucket_name} target_folder:#{target_folder}")
+
+      # Iterate over files in folder.
+      Dir[File.join(source_folder, '**/*')].reject { |fn| File.directory?(fn) }.each do |source_file|
+
+        # Convert to S3 path. Removing source parent folders and replacing with s3 output folder.
+        target_file = File.join(target_folder, source_file.gsub(source_folder, ''))
+        s3_copy_file_to_s3(bucket_name: bucket_name, source_file: source_file, target_file: target_file)
+      end
+    end
+
+    def s3_copy(source_folder:, target_folder:)
+      require 'open3'
+      exit_code = nil
+      error = nil
+      Open3.popen2e('aws', 's3', 'cp', source_folder, target_folder, '--recursive') do |stdin, stdout_stderr, wait_thread|
+        error = stdout_stderr.read
+        exit_code = wait_thread.value
+      end
+      if exit_code != 0
+        raise(error)
+      end
+
+      return exit_code
+    end
+
+    def output_hourly_data(model, output_folder, datapoint_id)
+      osm_path = File.join(output_folder, "run_dir/in.osm")
+      sql_path = File.join(output_folder, "run_dir/run/eplusout.sql")
+      csv_output = File.join(output_folder, "hourly.csv")
+
+      hours_of_year = []
+      d = Time.new(2006, 1, 1, 1)
+      (0...8760).each do |increment|
+        hours_of_year << (d + (60 * 60) * increment).strftime('%Y-%m-%d %H:%M')
+      end
+
+      array_of_hashes = []
+
+      # Find hourly outputs available for this datapoint.
+      query = "
+          SELECT ReportDataDictionaryIndex
+          FROM ReportDataDictionary
+          WHERE ReportingFrequency == 'Hourly'
+                                                         "
+      # Get hourly data for each output.
+      model.sqlFile.get.execAndReturnVectorOfInt(query).get.each do |rdd_index|
+
+        # Get Name
+        query = "
+          SELECT Name
+          FROM ReportDataDictionary
+          WHERE ReportDataDictionaryIndex == #{rdd_index}
+        "
+        name = model.sqlFile.get.execAndReturnFirstString(query).get
+
+        # Get KeyValue
+        query = "
+          SELECT KeyValue
+          FROM ReportDataDictionary
+          WHERE ReportDataDictionaryIndex == #{rdd_index}
+        "
+
+        # In some cases KeyValue has a value and sometimes it does not.  In some cases KeyValue is null.  If the command
+        # below is run and KeyValue is null then the command fails and returns an error.  The fix below assumes that if
+        # the command below fails it is because KeyValue is null.  In that case the "key_value" variable is set to a
+        # blank.
+        begin
+          key_value = model.sqlFile.get.execAndReturnFirstString(query).get
+        rescue StandardError => error
+          key_value = ""
+        end
+
+        # Get Units
+        query = "
+          SELECT Units
+          FROM ReportDataDictionary
+          WHERE ReportDataDictionaryIndex == #{rdd_index}
+        "
+        units = model.sqlFile.get.execAndReturnFirstString(query).get
+
+        # Get hourly data
+        query = "
+                      Select Value
+                      FROM ReportData
+                      WHERE
+                          ReportDataDictionaryIndex = #{rdd_index}
+        "
+        hourly_values = model.sqlFile.get.execAndReturnVectorOfDouble(query).get
+
+        hourly_hash = Hash[hours_of_year.zip(hourly_values)]
+
+        data_hash = { "datapoint_id": datapoint_id, "Name": name, "KeyValue": key_value, "Units": units }.merge(hourly_hash)
+        array_of_hashes << data_hash
+      end
+
+      CSV.open(csv_output, "wb") do |csv|
+        unless array_of_hashes.empty?
+          csv << array_of_hashes.first.keys # adds the attributes name on the first line
+          array_of_hashes.each do |hash|
+            csv << hash.values
+          end
+        end
+      end
+    end
+    #=====================================================================================================================
+    def output_timestep_data(model, output_folder,datapoint_id)
+      osm_path = File.join(output_folder, "run_dir/in.osm")
+      sql_path = File.join(output_folder, "run_dir/run/eplusout.sql")
+      csv_output = File.join(output_folder, "timestep.csv")
+      #===================================================================================================================
+      # Get number of timesteps from the model
+      number_of_timesteps_per_hour = model.getTimestep.numberOfTimestepsPerHour
+      #===================================================================================================================
+      # Calculate number of timesteps of the whole year
+      number_of_timesteps_of_year = 365 * 24 * number_of_timesteps_per_hour
+
+      timesteps_of_year = []
+      d = Time.new(2006, 1, 1, 0)
+      (0...number_of_timesteps_of_year).each do |increment|
+        timesteps_of_year << (d + (60 * 60 / number_of_timesteps_per_hour) * increment).strftime('%Y-%m-%d %H:%M')
+      end
+      timesteps_index = Array(0..number_of_timesteps_of_year-1)
+      #===================================================================================================================
+      # Create a hash with indices ('timesteps_index') as keys and timesteps as values ('timesteps_of_year')
+      # Note from Sara Gilani: 'I had to do this, as when I used 'timesteps_of_year' as keys to create all hashes below
+      # (e.g. timestep_hash_datapoint_id, timestep_hash_name, timestep_hash_key_value, ...),
+      # it did not create the right number of timesteps for the whole year'
+      timesteps_of_year_with_index = Hash[timesteps_index.zip(timesteps_of_year)]
+      #===================================================================================================================
+      array_of_hashes = []
+      array_of_hashes_transposed = []
+      #===================================================================================================================
+      # Find timestep outputs available for this datapoint.
+      query = "
+          SELECT ReportDataDictionaryIndex
+          FROM ReportDataDictionary
+          WHERE ReportingFrequency == 'Zone Timestep'
+          "
+      #===================================================================================================================
+      # Get timestep data for each output.
+      model.sqlFile.get.execAndReturnVectorOfInt(query).get.each do |rdd_index|
+        #=================================================================================================================
+        # Get Name
+        query = "
+          SELECT Name
+          FROM ReportDataDictionary
+          WHERE ReportDataDictionaryIndex == #{rdd_index}
+        "
+        name = model.sqlFile.get.execAndReturnFirstString(query).get
+        #=================================================================================================================
+        # Get KeyValue
+        query = "
+          SELECT KeyValue
+          FROM ReportDataDictionary
+          WHERE ReportDataDictionaryIndex == #{rdd_index}
+        "
+        #=================================================================================================================
+        # In some cases KeyValue has a value and sometimes it does not.  In some cases KeyValue is null.  If the command
+        # below is run and KeyValue is null then the command fails and returns an error.  The fix below assumes that if
+        # the command below fails it is because KeyValue is null.  In that case the "key_value" variable is set to a
+        # blank.
+        begin
+          key_value = model.sqlFile.get.execAndReturnFirstString(query).get
+        rescue StandardError => error
+          key_value = ""
+        end
+        #=================================================================================================================
+        # Get Units
+        query = "
+          SELECT Units
+          FROM ReportDataDictionary
+          WHERE ReportDataDictionaryIndex == #{rdd_index}
+        "
+        units = model.sqlFile.get.execAndReturnFirstString(query).get
+        #=================================================================================================================
+        # Get timestep data
+        query = "
+          Select VariableValue
+          FROM ReportVariableData
+          WHERE
+          ReportVariableDataDictionaryIndex = #{rdd_index}
+        "
+        timestep_values = model.sqlFile.get.execAndReturnVectorOfDouble(query).get
+
+        timestep_hash = Hash[timesteps_index.zip(timestep_values)]
+
+        data_hash = { "datapoint_id": datapoint_id, "Name": name, "KeyValue": key_value, "Units": units }.merge(timestep_hash)
+
+        array_of_hashes << data_hash
+
+        #=================================================================================================================
+        # Create a hash of data for having timesteps as rows instead of columns that is used for hourly outputs START
+        number_of_timesteps = timesteps_of_year.length()
+
+        array = [datapoint_id]
+        array_datapoint_id = array.zip(*[array]*number_of_timesteps).flatten
+        timestep_hash_datapoint_id = Hash[timesteps_index.zip(array_datapoint_id)]
+
+        array = [name]
+        array_name = array.zip(*[array]*number_of_timesteps).flatten
+        timestep_hash_name = Hash[timesteps_index.zip(array_name)]
+
+        array = [key_value]
+        array_key_value = array.zip(*[array]*number_of_timesteps).flatten
+        timestep_hash_key_value = Hash[timesteps_index.zip(array_key_value)]
+
+        array = [units]
+        array_units = array.zip(*[array]*number_of_timesteps).flatten
+        timestep_hash_units = Hash[timesteps_index.zip(array_units)]
+
+        timestep_hash_values = Hash[timesteps_index.zip(timestep_values)]
+
+        data_hash_transposed = Hash.new
+        timestep_hash_values.keys.each do |key|
+          data_hash_transposed[key] = [
+            timesteps_of_year_with_index[key],
+            timestep_hash_datapoint_id[key],
+            timestep_hash_name[key],
+            timestep_hash_key_value[key],
+            timestep_hash_units[key],
+            timestep_hash_values[key]
+          ]
+        end
+
+        array_of_hashes_transposed << data_hash_transposed
+        # Create a hash of data for having timesteps as rows instead of columns that is used for hourly outputs END
+        #=================================================================================================================
+
+      end #model.sqlFile.get.execAndReturnVectorOfInt(query).get.each do |rdd_index|
+      #===================================================================================================================
+      # # This csv file has timesteps as the first row
+      # CSV.open(csv_output, "wb") do |csv|
+      #   unless array_of_hashes.empty?
+      #     csv << array_of_hashes.first.keys # adds the attributes name on the first line
+      #     array_of_hashes.each do |hash|
+      #       csv << hash.values
+      #     end
+      #   end
+      # end
+      #===================================================================================================================
+      # This csv file has timesteps as the first column
+      header = ['Index', 'Timestep', 'datapoint_id', 'Name', 'KeyValue', 'Units', 'Value']
+      CSV.open(csv_output, "wb") do |csv|
+        csv << header
+        array_of_hashes_transposed.each do |hash|
+          hash.keys().each do |index|
+            row_index = index
+            row_timestep = hash[index][0]
+            row_datapoint_id = hash[index][1]
+            row_name = hash[index][2]
+            row_key_value = hash[index][3]
+            row_units = hash[index][4]
+            row_value = hash[index][5]
+
+            row = CSV::Row.new(header,[])
+            row['Index'] = row_index
+            row['Timestep'] = row_timestep
+            row['datapoint_id'] = row_datapoint_id
+            row['Name'] = row_name
+            row['KeyValue'] = row_key_value
+            row['Units'] = row_units
+            row['Value'] = row_value
+            csv << row
+          end
+        end
+      end
+      #===================================================================================================================
+    end #output_timestep_data
+    #=====================================================================================================================
+  end
+end