Data description for the Hotel Example in Section 5.2. 

The data is the outcome of processing the files described in Bodea et al. (2009). Our processing criteria is outlined in the main body of the paper.


The following parameters apply: n (number of products), N (number of customer types), T (number of periods).


For each Hotel Hi, i=1,...,5, the following text files are available:

ProductIDsHi.txt: List of records, with one row per hotel product (i.e., room type). Each record is of the form: (product label, description, average price).

PrefListsBuyUpHi.txt: Description of the customer types for the buy-up case. Matrix of (n+1) rows and N columns. Column k corresponds to the description 
of customer type k, so that position (1,k) has the prodcut label of the top choice for customer type k, position (2,k) has the second choice for 
customer type k, and so on. The type (i.e., column in the matrix) is an ordered list of up to n product labels, and is filled with zeroes up to row n+1.

PrefListsIndepHi.txt: Description of the customer types for the independent demand case. Structure similar to the file PrefListsBuyUpHi.txt.

AvailabilityHi.txt: Matrix with n rows and T columns. Position (j,t) has a 1 if product j is available in period t, and 0 otherwise.

TransactionsHi.txt: Row vector with T positions. Position t has the product label corresponding to the transaction observed in period t.
