colRowAnys

matrixStats: Benchmark report

colAnys() and rowAnys() benchmarks

This report benchmark the performance of colAnys() and rowAnys() against alternative methods.

Alternative methods

apply() + any()
colSums() > 0 or rowSums() > 0

Data

> rmatrix <- function(nrow, ncol, mode = c("logical", "double", "integer", "index"), range = c(-100, 
+     +100), naProb = 0) {
+     mode <- match.arg(mode)
+     n <- nrow * ncol
+     if (mode == "logical") {
+         X <- sample(c(FALSE, TRUE), size = n, replace = TRUE)
+     }     else if (mode == "index") {
+         X <- seq_len(n)
+         mode <- "integer"
+     }     else {
+         X <- runif(n, min = range[1], max = range[2])
+     }
+     storage.mode(X) <- mode
+     if (naProb > 0) 
+         X[sample(n, size = naProb * n)] <- NA
+     dim(X) <- c(nrow, ncol)
+     X
+ }
> rmatrices <- function(scale = 10, seed = 1, ...) {
+     set.seed(seed)
+     data <- list()
+     data[[1]] <- rmatrix(nrow = scale * 1, ncol = scale * 1, ...)
+     data[[2]] <- rmatrix(nrow = scale * 10, ncol = scale * 10, ...)
+     data[[3]] <- rmatrix(nrow = scale * 100, ncol = scale * 1, ...)
+     data[[4]] <- t(data[[3]])
+     data[[5]] <- rmatrix(nrow = scale * 10, ncol = scale * 100, ...)
+     data[[6]] <- t(data[[5]])
+     names(data) <- sapply(data, FUN = function(x) paste(dim(x), collapse = "x"))
+     data
+ }
> data <- rmatrices(mode = "logical")

Results

10x10 matrix

> X <- data[["10x10"]]
> gc()
           used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells   649471 34.7    1168576  62.5  1168576  62.5
Vcells 12125146 92.6   35610798 271.7 68120027 519.8
> colStats <- microbenchmark(colAnys = colAnys(X), `apply+any` = apply(X, MARGIN = 2L, FUN = any), 
+     `colSums > 0` = (colSums(X) > 0L), unit = "ms")
> X <- t(X)
> gc()
           used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells   648130 34.7    1168576  62.5  1168576  62.5
Vcells 12121512 92.5   35610798 271.7 68120027 519.8
> rowStats <- microbenchmark(rowAnys = rowAnys(X), `apply+any` = apply(X, MARGIN = 1L, FUN = any), 
+     `rowSums > 0` = (rowSums(X) > 0L), unit = "ms")

Table: Benchmarking of colAnys(), apply+any() and colSums > 0() on 10x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colAnys	0.0023	0.0035	0.0047	0.0050	0.0058	0.0192
3	colSums > 0	0.0058	0.0073	0.0088	0.0089	0.0100	0.0381
2	apply+any	0.0481	0.0520	0.0550	0.0533	0.0550	0.1190

	expr	min	lq	mean	median	uq	max
1	colAnys	1.000	1.000	1.000	1.000	1.000	1.00
3	colSums > 0	2.499	2.111	1.853	1.769	1.733	1.98
2	apply+any	20.823	14.994	11.594	10.653	9.532	6.18
Table: Benchmarking of rowAnys(), apply+any() and rowSums > 0() on 10x10 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowAnys	0.0027	0.0037	0.0049	0.0052	0.0058	0.0192
3	rowSums > 0	0.0058	0.0069	0.0088	0.0085	0.0100	0.0381
2	apply+any	0.0481	0.0520	0.0541	0.0535	0.0547	0.1193

	expr	min	lq	mean	median	uq	max
1	rowAnys	1.000	1.000	1.00	1.000	1.000	1.00
3	rowSums > 0	2.143	1.895	1.81	1.629	1.733	1.98
2	apply+any	17.856	14.207	11.08	10.294	9.466	6.20
Figure: Benchmarking of colAnys(), apply+any() and colSums > 0() on 10x10 data as well as rowAnys(), apply+any() and rowSums > 0() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colAnys() and rowAnys() on 10x10 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

expr	min	lq	mean	median	uq	max
colAnys	2.311	3.466	4.744	5.005	5.775	19.25
rowAnys	2.695	3.658	4.886	5.198	5.775	19.25

expr	min	lq	mean	median	uq	max
colAnys	1.000	1.000	1.00	1.000	1	1
rowAnys	1.166	1.055	1.03	1.039	1	1
Figure: Benchmarking of colAnys() and rowAnys() on 10x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x100 matrix

> X <- data[["100x100"]]
> gc()
           used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells   648192 34.7    1168576  62.5  1168576  62.5
Vcells 12122694 92.5   35610798 271.7 68120027 519.8
> colStats <- microbenchmark(colAnys = colAnys(X), `apply+any` = apply(X, MARGIN = 2L, FUN = any), 
+     `colSums > 0` = (colSums(X) > 0L), unit = "ms")
> X <- t(X)
> gc()
           used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells   648186 34.7    1168576  62.5  1168576  62.5
Vcells 12127737 92.6   35610798 271.7 68120027 519.8
> rowStats <- microbenchmark(rowAnys = rowAnys(X), `apply+any` = apply(X, MARGIN = 1L, FUN = any), 
+     `rowSums > 0` = (rowSums(X) > 0L), unit = "ms")

Table: Benchmarking of colAnys(), apply+any() and colSums > 0() on 100x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colAnys	0.0065	0.0092	0.0127	0.0123	0.0150	0.0293
3	colSums > 0	0.0192	0.0358	0.0397	0.0397	0.0429	0.1174
2	apply+any	0.4273	0.6562	0.6697	0.6929	0.7195	0.8284

	expr	min	lq	mean	median	uq	max
1	colAnys	1.000	1.000	1.000	1.000	1.000	1.000
3	colSums > 0	2.941	3.875	3.117	3.219	2.859	4.013
2	apply+any	65.287	71.012	52.622	56.246	47.919	28.315
Table: Benchmarking of rowAnys(), apply+any() and rowSums > 0() on 100x100 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowAnys	0.0150	0.0169	0.0191	0.0189	0.0200	0.0354
3	rowSums > 0	0.0385	0.0404	0.0439	0.0418	0.0450	0.0670
2	apply+any	0.3538	0.3657	0.4183	0.3717	0.4333	0.7245

	expr	min	lq	mean	median	uq	max
1	rowAnys	1.000	1.000	1.000	1.000	1.00	1.000
3	rowSums > 0	2.564	2.386	2.293	2.214	2.25	1.891
2	apply+any	23.563	21.590	21.870	19.704	21.64	20.456
Figure: Benchmarking of colAnys(), apply+any() and colSums > 0() on 100x100 data as well as rowAnys(), apply+any() and rowSums > 0() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colAnys() and rowAnys() on 100x100 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

expr	min	lq	mean	median	uq	max
colAnys	6.545	9.24	12.73	12.32	15.01	29.26
rowAnys	15.014	16.94	19.13	18.86	20.02	35.42

expr	min	lq	mean	median	uq	max
colAnys	1.000	1.000	1.000	1.000	1.000	1.00
rowAnys	2.294	1.833	1.503	1.531	1.333	1.21
Figure: Benchmarking of colAnys() and rowAnys() on 100x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x10 matrix

> X <- data[["1000x10"]]
> gc()
           used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells   648232 34.7    1168576  62.5  1168576  62.5
Vcells 12122935 92.5   35610798 271.7 68120027 519.8
> colStats <- microbenchmark(colAnys = colAnys(X), `apply+any` = apply(X, MARGIN = 2L, FUN = any), 
+     `colSums > 0` = (colSums(X) > 0L), unit = "ms")
> X <- t(X)
> gc()
           used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells   648226 34.7    1168576  62.5  1168576  62.5
Vcells 12127978 92.6   35610798 271.7 68120027 519.8
> rowStats <- microbenchmark(rowAnys = rowAnys(X), `apply+any` = apply(X, MARGIN = 1L, FUN = any), 
+     `rowSums > 0` = (rowSums(X) > 0L), unit = "ms")

Table: Benchmarking of colAnys(), apply+any() and colSums > 0() on 1000x10 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colAnys	0.0023	0.0039	0.0059	0.0062	0.0069	0.0558
3	colSums > 0	0.0162	0.0181	0.0202	0.0212	0.0219	0.0362
2	apply+any	0.2202	0.2229	0.2408	0.2256	0.2371	0.3472

	expr	min	lq	mean	median	uq	max
1	colAnys	1.000	1.000	1.000	1.000	1.000	1.0000
3	colSums > 0	6.997	4.699	3.414	3.437	3.166	0.6483
2	apply+any	95.281	57.886	40.711	36.621	34.218	6.2205
Table: Benchmarking of rowAnys(), apply+any() and rowSums > 0() on 1000x10 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowAnys	0.0200	0.0208	0.0229	0.0229	0.0237	0.0466
3	rowSums > 0	0.0385	0.0406	0.0428	0.0433	0.0447	0.0554
2	apply+any	0.2194	0.2227	0.2426	0.2260	0.2379	0.3757

	expr	min	lq	mean	median	uq	max
1	rowAnys	1.000	1.000	1.000	1.000	1.000	1.000
3	rowSums > 0	1.923	1.954	1.871	1.891	1.886	1.190
2	apply+any	10.961	10.712	10.603	9.865	10.048	8.066
Figure: Benchmarking of colAnys(), apply+any() and colSums > 0() on 1000x10 data as well as rowAnys(), apply+any() and rowSums > 0() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colAnys() and rowAnys() on 1000x10 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

expr	min	lq	mean	median	uq	max
colAnys	2.311	3.85	5.914	6.16	6.93	55.82
rowAnys	20.018	20.79	22.879	22.91	23.68	46.58

expr	min	lq	mean	median	uq	max
colAnys	1.000	1.000	1.000	1.000	1.000	1.0000
rowAnys	8.662	5.399	3.869	3.719	3.417	0.8345
Figure: Benchmarking of colAnys() and rowAnys() on 1000x10 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

10x1000 matrix

> X <- data[["10x1000"]]
> gc()
           used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells   648262 34.7    1168576  62.5  1168576  62.5
Vcells 12123503 92.5   35610798 271.7 68120027 519.8
> colStats <- microbenchmark(colAnys = colAnys(X), `apply+any` = apply(X, MARGIN = 2L, FUN = any), 
+     `colSums > 0` = (colSums(X) > 0L), unit = "ms")
> X <- t(X)
> gc()
           used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells   648256 34.7    1168576  62.5  1168576  62.5
Vcells 12128546 92.6   35610798 271.7 68120027 519.8
> rowStats <- microbenchmark(rowAnys = rowAnys(X), `apply+any` = apply(X, MARGIN = 1L, FUN = any), 
+     `rowSums > 0` = (rowSums(X) > 0L), unit = "ms")

Table: Benchmarking of colAnys(), apply+any() and colSums > 0() on 10x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colAnys	0.0154	0.0179	0.0268	0.0252	0.0329	0.0543
3	colSums > 0	0.0266	0.0300	0.0428	0.0381	0.0524	0.1012
2	apply+any	1.6950	1.7841	2.3478	1.8484	2.6893	6.7325

	expr	min	lq	mean	median	uq	max
1	colAnys	1.000	1.000	1.000	1.000	1.000	1.000
3	colSums > 0	1.725	1.677	1.599	1.511	1.591	1.865
2	apply+any	110.069	99.663	87.677	73.302	81.705	124.034
Table: Benchmarking of rowAnys(), apply+any() and rowSums > 0() on 10x1000 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowAnys	0.0339	0.0358	0.0430	0.0393	0.0491	0.0731
3	rowSums > 0	0.0474	0.0493	0.0601	0.0549	0.0654	0.1382
2	apply+any	1.6965	1.7806	2.2477	1.8491	2.1284	4.5579

	expr	min	lq	mean	median	uq	max
1	rowAnys	1.000	1.000	1.000	1.000	1.000	1.000
3	rowSums > 0	1.398	1.376	1.398	1.397	1.333	1.889
2	apply+any	50.078	49.736	52.324	47.092	43.364	62.314
Figure: Benchmarking of colAnys(), apply+any() and colSums > 0() on 10x1000 data as well as rowAnys(), apply+any() and rowSums > 0() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colAnys() and rowAnys() on 10x1000 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

expr	min	lq	mean	median	uq	max
colAnys	15.40	17.9	26.78	25.22	32.91	54.28
rowAnys	33.88	35.8	42.96	39.27	49.08	73.14

expr	min	lq	mean	median	uq	max
colAnys	1.0	1	1.000	1.000	1.000	1.000
rowAnys	2.2	2	1.604	1.557	1.491	1.347
Figure: Benchmarking of colAnys() and rowAnys() on 10x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

100x1000 matrix

> X <- data[["100x1000"]]
> gc()
           used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells   648306 34.7    1168576  62.5  1168576  62.5
Vcells 12123901 92.5   35610798 271.7 68120027 519.8
> colStats <- microbenchmark(colAnys = colAnys(X), `apply+any` = apply(X, MARGIN = 2L, FUN = any), 
+     `colSums > 0` = (colSums(X) > 0L), unit = "ms")
> X <- t(X)
> gc()
           used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells   648300 34.7    1168576  62.5  1168576  62.5
Vcells 12173944 92.9   35610798 271.7 68120027 519.8
> rowStats <- microbenchmark(rowAnys = rowAnys(X), `apply+any` = apply(X, MARGIN = 1L, FUN = any), 
+     `rowSums > 0` = (rowSums(X) > 0L), unit = "ms")

Table: Benchmarking of colAnys(), apply+any() and colSums > 0() on 100x1000 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colAnys	0.0177	0.0223	0.0472	0.0395	0.0679	0.127
3	colSums > 0	0.1290	0.1320	0.1983	0.1825	0.2520	0.390
2	apply+any	3.3456	3.6267	5.0826	4.0397	6.0815	22.921

	expr	min	lq	mean	median	uq	max
1	colAnys	1.000	1.000	1.000	1.000	1.000	1.00
3	colSums > 0	7.282	5.914	4.203	4.624	3.708	3.07
2	apply+any	188.923	162.426	107.736	102.379	89.506	180.43
Table: Benchmarking of rowAnys(), apply+any() and rowSums > 0() on 100x1000 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowAnys	0.1240	0.1293	0.1549	0.1519	0.1728	0.2221
3	rowSums > 0	0.3345	0.3703	0.4350	0.4177	0.5139	0.6232
2	apply+any	3.3861	4.3966	5.8292	5.3917	6.6093	22.0532

	expr	min	lq	mean	median	uq	max
1	rowAnys	1.000	1.000	1.000	1.00	1.000	1.000
3	rowSums > 0	2.699	2.863	2.809	2.75	2.973	2.806
2	apply+any	27.317	33.991	37.639	35.50	38.238	99.286
Figure: Benchmarking of colAnys(), apply+any() and colSums > 0() on 100x1000 data as well as rowAnys(), apply+any() and rowSums > 0() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colAnys() and rowAnys() on 100x1000 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

expr	min	lq	mean	median	uq	max
colAnys	17.71	22.33	47.18	39.46	67.94	127.0
rowAnys	123.96	129.34	154.87	151.87	172.84	222.1

expr	min	lq	mean	median	uq	max
colAnys	1	1.000	1.000	1.000	1.000	1.000
rowAnys	7	5.793	3.283	3.849	2.544	1.748
Figure: Benchmarking of colAnys() and rowAnys() on 100x1000 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

1000x100 matrix

> X <- data[["1000x100"]]
> gc()
           used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells   648337 34.7    1168576  62.5  1168576  62.5
Vcells 12124353 92.6   35610798 271.7 68120027 519.8
> colStats <- microbenchmark(colAnys = colAnys(X), `apply+any` = apply(X, MARGIN = 2L, FUN = any), 
+     `colSums > 0` = (colSums(X) > 0L), unit = "ms")
> X <- t(X)
> gc()
           used (Mb) gc trigger  (Mb) max used  (Mb)
Ncells   648331 34.7    1168576  62.5  1168576  62.5
Vcells 12174396 92.9   35610798 271.7 68120027 519.8
> rowStats <- microbenchmark(rowAnys = rowAnys(X), `apply+any` = apply(X, MARGIN = 1L, FUN = any), 
+     `rowSums > 0` = (rowSums(X) > 0L), unit = "ms")

Table: Benchmarking of colAnys(), apply+any() and colSums > 0() on 1000x100 data. The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	colAnys	0.0050	0.0083	0.0208	0.0204	0.0320	0.0589
3	colSums > 0	0.1136	0.1164	0.1682	0.1621	0.2179	0.2768
2	apply+any	1.8809	1.9681	2.5582	2.6518	2.8833	3.8815

	expr	min	lq	mean	median	uq	max
1	colAnys	1.00	1.00	1.000	1.000	1.000	1.000
3	colSums > 0	22.69	14.07	8.079	7.943	6.819	4.699
2	apply+any	375.80	237.76	122.897	129.963	90.239	65.901
Table: Benchmarking of rowAnys(), apply+any() and rowSums > 0() on 1000x100 data (transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

	expr	min	lq	mean	median	uq	max
1	rowAnys	0.1163	0.1201	0.1380	0.1334	0.1386	0.5054
3	rowSums > 0	0.3361	0.3397	0.3917	0.3644	0.4619	0.6113
2	apply+any	1.9367	2.3575	2.8386	2.7830	3.0758	4.3946

	expr	min	lq	mean	median	uq	max
1	rowAnys	1.000	1.000	1.000	1.000	1.000	1.000
3	rowSums > 0	2.891	2.829	2.839	2.732	3.333	1.209
2	apply+any	16.659	19.628	20.576	20.864	22.194	8.695
Figure: Benchmarking of colAnys(), apply+any() and colSums > 0() on 1000x100 data as well as rowAnys(), apply+any() and rowSums > 0() on the same data transposed. Outliers are displayed as crosses. Times are in milliseconds.

Table: Benchmarking of colAnys() and rowAnys() on 1000x100 data (original and transposed). The top panel shows times in milliseconds and the bottom panel shows relative times.

expr	min	lq	mean	median	uq	max
colAnys	5.005	8.277	20.82	20.4	31.95	58.9
rowAnys	116.258	120.106	137.96	133.4	138.58	505.4

expr	min	lq	mean	median	uq	max
colAnys	1.00	1.00	1.000	1.000	1.000	1.000
rowAnys	23.23	14.51	6.628	6.537	4.337	8.582
Figure: Benchmarking of colAnys() and rowAnys() on 1000x100 data (original and transposed). Outliers are displayed as crosses. Times are in milliseconds.

Appendix

Session information

R Under development (unstable) (2015-02-27 r67909)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 7 x64 (build 7601) Service Pack 1

locale:
[1] LC_COLLATE=English_United States.1252 
[2] LC_CTYPE=English_United States.1252   
[3] LC_MONETARY=English_United States.1252
[4] LC_NUMERIC=C                          
[5] LC_TIME=English_United States.1252    

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] markdown_0.7.7          microbenchmark_1.4-2    matrixStats_0.14.0-9000
[4] ggplot2_1.0.0           knitr_1.9.3             R.devices_2.13.0       
[7] R.utils_2.0.0           R.oo_1.19.0             R.methodsS3_1.7.0      

loaded via a namespace (and not attached):
 [1] Rcpp_0.11.4         splines_3.2.0       MASS_7.3-39        
 [4] munsell_0.4.2       lattice_0.20-30     colorspace_1.2-4   
 [7] R.cache_0.11.1-9000 multcomp_1.3-9      stringr_0.6.2      
[10] plyr_1.8.1          tools_3.2.0         grid_3.2.0         
[13] gtable_0.1.2        TH.data_1.0-6       survival_2.38-1    
[16] digest_0.6.8        R.rsp_0.20.0        reshape2_1.4.1     
[19] formatR_1.0.3       base64enc_0.1-3     mime_0.2.1         
[22] evaluate_0.5.7      labeling_0.3        sandwich_2.3-2     
[25] scales_0.2.4        mvtnorm_1.0-2       zoo_1.7-12         
[28] Cairo_1.5-6         proto_0.3-10

Total processing time was 17.31 secs.

Reproducibility

To reproduce this report, do:

html <- matrixStats:::benchmark('colAnys')

Copyright Henrik Bengtsson. Last updated on 2015-03-02 16:58:09 (-0800 UTC). Powered by RSP.

[Benchmark reports](Benchmark reports)

colRowAnys

colAnys() and rowAnys() benchmarks

Alternative methods

Data

Results

10x10 matrix

100x100 matrix

1000x10 matrix

10x1000 matrix

100x1000 matrix

1000x100 matrix

Appendix

Session information

Reproducibility

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Clone this wiki locally