Changeset 178
Legend:
 Unmodified
 Added
 Removed

trunk/src/NNI.h
r177 r178 33 33 /// corrections, whereas a 0 means that the data should be replaced 34 34 /// by an imputed value. Naturally, the weight matrix can also be 35 /// continuo s where values between 0 and 1 defines how certain a35 /// continuous where values between 0 and 1 defines how certain a 36 36 /// data element is. 37 37 /// … … 42 42 /// Implementation issues 43 43 /// 44 /// The current implementation treats rows where all data are tagge t45 /// are completely uncerta tin, i.e. all weights are zero, by44 /// The current implementation treats rows where all data are tagged 45 /// are completely uncertain, i.e. all weights are zero, by 46 46 /// ignoring these lines in nearest neighbourhood 47 47 /// calculations. Importantly, this type of data are not changed … … 54 54 /// number. Identical rows in this context are basically a 55 55 /// comparison between elements with nonzero uncertainty weights 56 /// only, and all these elem nts are equal. Zero weight elemenets are56 /// only, and all these elements are equal. Zero weight elements are 57 57 /// not used in the comparison since these are considered as 58 58 /// non/sense values. … … 61 61 { 62 62 public: 63 64 /// 63 65 /// Base constructor for the nearest neighbour imputation 64 66 /// algorithms. 67 /// 65 68 NNI(const gslapi::matrix& matrix,const gslapi::matrix& weight, 66 69 const u_int neighbours); … … 68 71 virtual ~NNI(void) {}; 69 72 73 /// 70 74 /// @return A const reference to the modified data. 75 /// 71 76 const gslapi::matrix& imputed_data(void) const { return imputed_data_; } 72 77 
trunk/src/WeNNI.h
r177 r178 16 16 17 17 /// 18 /// WeNNI is a contin ous weights generalization of the (binary19 /// weights) kNNI algorithm presented by Troyanska et al. A18 /// WeNNI is a continuous weights generalization of the (binary 19 /// weights) kNNI algorithm presented by Troyanskaya et al. A 20 20 /// reference to this paper is found in the NNI document referred to 21 21 /// in the NNI class documentation. The NNI document also describes … … 27 27 { 28 28 public: 29 /// Perform WeNNI on data in \a matrix with contin ious uncertainty29 /// Perform WeNNI on data in \a matrix with continuous uncertainty 30 30 /// weights in \a weight using \a neighbours for the new impute 31 31 /// value. 
trunk/src/kNNI.h
r177 r178 14 14 /// 15 15 /// kNNI is the binary weight implementation of NNI. This follows 16 /// the work done by Troyanska et al. cited in the NNI document16 /// the work done by Troyanskaya et al. cited in the NNI document 17 17 /// referred to in the NNI class documentation. 18 18 ///
Note: See TracChangeset
for help on using the changeset viewer.