Reading NIPS 360 FFS Data Files

NIPS 360 FFS Data Files is a collection of logical records (see nips.logical_records.LogicalRecord) that are parsed into various kind of records holding the information contained in the data file.

The simplest way to load a data file to a data set (nips.DataSet). The data set consists of all the parsed records in the data file:

import nips

with open("FILE.NIPS", "rb") as nip_file:
    data_set = nips.DataSet(nips_file)

Data Set

class nips.DataSet(nips_file)

NIPS Data Set

A Python representation of a NIPS Data Set with all data loaded into memory.

This class is useful for loading an entire NIPS Data Set for interactive or programmatic inspection.

Note that the entire data set is loaded into memory. To process records without loading them all into memory use nips.logical_record.read() and the nips.fft.FileFormatTable class.

Parameters:

nips_file (file-like object) – File from where data is loaded.

fft

File Format Table of the data set.

Type:

FileFormatTable

fixed_set

Fixed Set.

Type:

list(DataFileRecord))

periodic_sets

Periodic Sets. Note that the list is 0-indexed. So the Periodic Set 1 is at periodic_sets[0].

Type:

list(list(DataFileRecord)))

orphans

Data file records that could not be assigned to a set. Maybe an invalid set id?

Type:

list(DataFileRecord)

record_type_count

A dictionary from record type code to count of records in NIPS file.

Type:

dict