Sanghamitra Das and Ramprasad Sengupta, "Projection Pursuit Regression
and Disaggregate Productivity Effects: The Case of the Indian Blast 
Furnaces", Journal of Applied Econometrics, Vol. 19, No. 3, 2004, pp.
397-418.

The dataset used in this paper has 825 observations (in rows) and the
nonconfidential part of the data has 13 variables.  (The labour and
training information in the paper is confidential and hence not included
here.) The name of the file, which is an ASCII file in DOS format, is
allvar3_jae.dat. It is zipped in the file ds-data.zip.

The 13 variables are ordered as follows.

1. Furnace ID (1-16)
2. Furnace productivity/day (tons/cubic meter/day)
3. Furnace productivity/utilized day (tons/cubic meter/utilized day)
4. Ash %
5. Micum %
6. Alumina-silica ratio
7. Sinter proportion in ore
8. Relining age (in months)
9. age (in months)
10. size (cubic metres)
11. planned delays (days/month)
12. cumulative planned delays (days)
13. unplanned delays (days/month)

Data Source: Operations Statistics published each year by the 
Steel Authority of India limited (SAIL in short) for each of its
plants.


Sanghamitra Das
Professor of Economics
Indian Statistical Institute
Qutub Institutional Area
New Delhi 110016
fax:+(91)-11-26856779
tel:+(91)-11-26564789
                 5050/ext. 205
             26514594 for messages