File: 
http://www.isip.piconepress.com/courses/temple/ece_8527/resources/data/set_10

This directory contains four versions of the data:
 
 - 00_float_32: the original continuous data
 
 - 01_uint_32: set 10 in which 00* was converted to bits using
   a 32-bit unsigned integer (same value, just bits instead of floats)

 - 02_lin_32: linear quantization to 32 bits in range of (-1.4,1.4).
	      
 - 03_lin_16: linear quantization to 16 bits in range of (-1.4,1.4).

 - 04_lin_16_nodups: linear quantization to 16 bits in range of (-1.4,1.4).
   Any point that appears with more than one label was modified so that
   the point only appears once and does not have ambiguous labels.

These sets were created to support some preliminary experiments in
quantum computing.

Each directory contains files named as follows:

nedc_000_[1]: ls -1 00_float_32/data/*.csv
00_float_32/data/00_dev_03.csv             <* original floats
00_float_32/data/00_eval_03.csv            <* original floats
00_float_32/data/00_train_03.csv           <* original floats

nedc_000_[1]: ls -1 01_uint_32/data
01_dev_03_bin_pairs.csv                    <* binary data as (x,y)
01_dev_03_bin.csv                          <* binary data as 1 feature
01_dev_03_sep.csv                          <* binary/separated

01_eval_03.csv                             <* quantized data as (x,y)
01_eval_03_bin_pairs.csv                   <* binary data as (x,y)
01_eval_03_bin.csv                         <* binary data as 1 feature
01_eval_03_sep.csv                         <* binary/separated

01_train_03.csv                            <* quantized data as (x,y)
01_train_03_bin_pairs.csv                  <* binary data as (x,y)
01_train_03_bin.csv                        <* binary data as 1 feature
01_train_03_sep.csv                        <* binary/separated

nedc_000_[1]: ls -1 02_lin_32/data
02_dev_03.csv                              <* quantized data as (x,y)
02_dev_03_bin_pairs.csv                    <* binary data as (x,y)
02_dev_03_bin.csv                          <* binary data as 1 feature
02_dev_03_sep.csv                          <* binary/separated

02_eval_03.csv                             <* quantized data as (x,y)
02_eval_03_bin_pairs.csv                   <* binary data as (x,y)
02_eval_03_bin.csv                         <* binary data as 1 feature
02_eval_03_sep.csv                         <* binary/separated

02_train_03.csv                            <* quantized data as (x,y)
02_train_03_bin_pairs.csv                  <* binary data as (x,y)
02_train_03_bin.csv                        <* binary data as 1 feature
02_train_03_sep.csv                        <* binary/separated

nedc_000_[1]: ls -1 03_lin_16/data
03_dev_03.csv                              <* quantized data as (x,y)
03_dev_03_bin_pairs.csv                    <* binary data as (x,y)
03_dev_03_bin.csv                          <* binary data as 1 feature
03_dev_03_sep.csv                          <* binary/separated

03_eval_03.csv                             <* quantized data as (x,y)
03_eval_03_bin_pairs.csv                    <* binary data as (x,y)
03_eval_03_bin.csv                         <* binary data as 1 feature
03_eval_03_sep.csv                         <* binary/separated

03_train_03.csv                            <* quantized data as (x,y)
03_train_03_bin_pairs.csv                  <* binary data as (x,y)
03_train_03_bin.csv                        <* binary data as 1 feature
03_train_03_sep.csv                        <* binary/separated

nedc_000_[1]: ls -1 04_lin_16_nodups/data
04_train_03_float.csv                      <* original nodups data
04_train_03.csv                            <* quantized data as (x,y)
04_train_03_bin_pairs.csv                  <* binary data as (x,y)
04_train_03_bin.csv                        <* binary data as 1 feature
04_train_03_sep.csv                        <* binary/separated

quantized as (x,y) pairs means you feed the data to the system as the
equivalent quantized floating point values. The attached spreadsheet
demonstrates the quantization process.