sample size k build changed, tests added

.Rproj.user/E3DB6272/pcs/files-pane.pper

@@ -9,5 +9,5 @@
             "ascending": false
         }
     ],
-    "path": "~/git/nauka/ncn-foreigners/software/blocking/R"
+    "path": "~/git/nauka/ncn-foreigners/software/blocking"
 }

.Rproj.user/E3DB6272/pcs/source-pane.pper

@@ -1,4 +1,4 @@
 {
-    "activeTab": 0,
+    "activeTab": 2,
     "activeTabSourceWindow0": 0
 }

.Rproj.user/shared/notebooks/paths

@@ -9,6 +9,7 @@
 /Users/berenz/git/nauka/ncn-foreigners/software/blocking/R/method_annoy.R="B0938ADD"
 /Users/berenz/git/nauka/ncn-foreigners/software/blocking/R/method_hnsw.R="C19508EB"
 /Users/berenz/git/nauka/ncn-foreigners/software/blocking/R/method_mlpack.R="9402F0E3"
+/Users/berenz/git/nauka/ncn-foreigners/software/blocking/R/method_nnd.R="E5F797EE"
 /Users/berenz/git/nauka/ncn-foreigners/software/blocking/R/methods.R="081419BC"
 /Users/berenz/git/nauka/ncn-foreigners/software/blocking/R/reclin2_pair_ann.R="D089A6FC"
 /Users/berenz/git/nauka/ncn-foreigners/software/blocking/README.Rmd="2B1049F0"

NEWS.md

@@ -10,3 +10,4 @@
 8. first vignette added.
 9. evaluation with standard metrics (recall, fpr etc) added, works with vector for deduplication.
 10. added saving index for hnsw and annoy
+11. `rnndescend` support added.

R/method_nnd.R

@@ -27,7 +27,7 @@ method_nnd <- function(x,
                        control) {
 
   l_ind <- rnndescent::rnnd_build(data = x,
-                                  k = control$nnd$k_build,
+                                  k = if (nrow(x) < control$nnd$k_build) nrow(x) else control$nnd$k_build,
                                   metric = distance,
                                   verbose = verbose,
                                   n_threads = n_threads,

README.Rmd

@@ -24,6 +24,7 @@ A small package used to block records for data deduplication and record linkage
 
 Currently supports the following R packages that binds to specific ANN algorithms
 
++ [rnndescent](https://cran.r-project.org/package=rnndescent) (default),
 + [RcppHNSW](https://cran.r-project.org/package=RcppHNSW), 
 + [RcppAnnoy](https://cran.r-project.org/package=RcppAnnoy), 
 + [mlpack](https://cran.r-project.org/package=RcppAnnoy) (see `mlpack::lsh` and `mlpack::knn`).

README.md

@@ -19,6 +19,7 @@ and graphs (via `igraph`).
 Currently supports the following R packages that binds to specific ANN
 algorithms
 
+- [rnndescent](https://cran.r-project.org/package=rnndescent) (default),
 - [RcppHNSW](https://cran.r-project.org/package=RcppHNSW),
 - [RcppAnnoy](https://cran.r-project.org/package=RcppAnnoy),
 - [mlpack](https://cran.r-project.org/package=RcppAnnoy) (see
@@ -93,13 +94,10 @@ Deduplication using blocking
 
 ``` r
 blocking_result <- blocking(x = df_example$txt)
-#> 'as(<dgTMatrix>, "dgCMatrix")' is deprecated.
-#> Use 'as(., "CsparseMatrix")' instead.
-#> See help("Deprecated") and help("Matrix-deprecated").
 ## data frame with indices and block 
 blocking_result
 #> ========================================================
-#> Blocking based on the hnsw method.
+#> Blocking based on the nnd method.
 #> Number of blocks: 2.
 #> Number of columns used for blocking: 28.
 #> Reduction ratio: 0.5714.
@@ -113,16 +111,16 @@ Table with blocking
 
 ``` r
 blocking_result$result
-#>        x     y block
-#>    <int> <int> <num>
-#> 1:     1     2     1
-#> 2:     2     1     1
-#> 3:     2     3     1
-#> 4:     2     4     1
-#> 5:     5     6     2
-#> 6:     5     7     2
-#> 7:     5     8     2
-#> 8:     6     5     2
+#>        x     y block       dist
+#>    <int> <int> <num>      <num>
+#> 1:     1     2     1 0.10000005
+#> 2:     1     3     1 0.14188367
+#> 3:     1     4     1 0.28286284
+#> 4:     2     1     1 0.10000005
+#> 5:     5     6     2 0.08333336
+#> 6:     5     7     2 0.13397458
+#> 7:     5     8     2 0.27831215
+#> 8:     6     5     2 0.08333336
 ```
 
 Deduplication followed by the `reclin2` package

inst/tinytest/test_reclin2.R

@@ -1,7 +1,7 @@
 source("test_data.R")
 
 expect_silent(
-  pair_ann(x = df_example, on = "txt")
+  pair_ann(x = df_example, on = "txt", ann = "hnsw")
 )
 
 expect_equal(