An error in measure writing: "bad index"?

edit

Dear the members of unmet-hours,

Currently, I'm writing measure to output the surface inside conduction in HAMT simulation by outputting variables surface heat balances in a table. (Related to this question)

I already modified "Add Output Variables" and succeeded. After that, currently I am trying to output those variables in one table by modifying "Export Variable To CSV".

When applying the code, the error message "bad index" comes up again and again in the line "value = values[kk][i]" to substitute an element of array to a variable... But in the error check sentence I added, index worked. I would like to know what the cause of this is.

Please check the code below (I am modifying "Export Varible To CSV"'s L.91 and later mainly).

variable_names = sqlFile.availableVariableNames(ann_env_pd, reporting_frequency)

heles = ["Surface Inside Face Convection Heat Gain Rate per Area","Surface Inside Face Net Surface Thermal Radiation Heat Gain Rate per Area","Surface Inside Face Solar Radiation Heat Gain Rate per Area","Surface Inside Face Lights Radiation Heat Gain Rate","Surface Inside Face Internal Gains Radiation Heat Gain Rate per Area","Surface Inside Face System Radiation Heat Gain Rate per Area"]

csv_array = []
cnt=0

for e in 0..5
  hele=heles[e]
  if !variable_names.include? "#{hele}"   
    runner.registerError("#{hele} is not in sqlFile.  Please add an AddOutputVariable reporting measure with this variable and run again.")
  else      
    #headers = ["#{reporting_frequency}"]
    headers = []
    output_timeseries = {}
    key_values = sqlFile.availableKeyValues(ann_env_pd, reporting_frequency, hele.to_s)
    if key_values.size == 0
       runner.registerError("Timeseries for #{hele} did not have any key values. No timeseries available.")
    end

    key_values.each do |key_value|
      timeseries = sqlFile.timeSeries(ann_env_pd, reporting_frequency, hele.to_s, key_value.to_s)
      if !timeseries.empty?
        timeseries = timeseries.get
        units = timeseries.units
        headers << "#{key_value.to_s}:#{hele.to_s}[#{units}]"
        output_timeseries[headers[-1]] = timeseries
      else 
        runner.registerWarning("Timeseries for #{key_value} #{hele} is empty.")
      end     
    end
    earray = []
    earray << headers
    date_times = ["Time"] + output_timeseries[output_timeseries.keys[0]].dateTimes

    values = {}
    for key in output_timeseries.keys
      kk=headers.find_index(key)
      values[kk] = output_timeseries[key].values
    end

    num_times = date_times.size - 1

    if e==0
      csv_array << date_times
    end

    for i in 0..num_times
      row = []

      for key in headers
        kk=headers.find_index(key)
        runner.registerInfo("#{key}: #{values[kk][i]} & Next: #{values[kk][i+1]}") if i==0 
        value = values[kk][i] #L.137error happens here as "bad index"
        row << value
      end
      earray << row
    end


    csv_array << earray

  end
end

csv_array=csv_array.transpose     

File.open("./report_condHFinHAMT_#{reporting_frequency.delete(' ')}.csv", 'wb') do |file|
  csv_array.each do |elem|
    file.puts elem.join(',')
  end
end

runner.registerInfo("Output file written to #{File.expand_path('.')}")


# close the sql file
sqlFile.close()

*Actually, usually I use the fortran for main. I have a poor knowledge about Ruby coding...

Add (July 17th 2018)

Whole messages shown in OS are below:

Line 137 is "value = values[kk][i]" as commented in the code above.