outliertree 0.3.1 → 0.4.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +4 -4
- data/CHANGELOG.md +9 -0
- data/README.md +1 -1
- data/lib/outliertree/result.rb +2 -2
- data/lib/outliertree/version.rb +1 -1
- data/lib/outliertree.rb +4 -4
- data/vendor/outliertree/src/clusters.cpp +41 -27
- data/vendor/outliertree/src/fit_model.cpp +21 -11
- data/vendor/outliertree/src/outlier_tree.hpp +2 -1
- metadata +6 -10
checksums.yaml
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
---
|
|
2
2
|
SHA256:
|
|
3
|
-
metadata.gz:
|
|
4
|
-
data.tar.gz:
|
|
3
|
+
metadata.gz: b2c75c112439d30795757595ab3d34de3fbe80049f1b03e7168f0d1eed405417
|
|
4
|
+
data.tar.gz: e08ddb70bdf7b1be287b2986fc9364aa554db465fc2ef11b193b8f2adbd7af19
|
|
5
5
|
SHA512:
|
|
6
|
-
metadata.gz:
|
|
7
|
-
data.tar.gz:
|
|
6
|
+
metadata.gz: 8bf35fb1fad7023c10b5b7514281d6db9c57faacfa8fcbaaa5e39a2aefb43508f5ed1c9c4cac7dd7e96052fac5e0c6036de201223ec3df7241051426e67a349c
|
|
7
|
+
data.tar.gz: 36d0ffa632d62be0f2ebe73bb7cace6b4f31eddf6202f26afe8250f27a8bbf25606be9ca8e9b5c89a2596b3acdc0c7124cf039d7b5e9557902a08ae233416293
|
data/CHANGELOG.md
CHANGED
data/README.md
CHANGED
|
@@ -10,7 +10,7 @@ Price (2.50) looks low given Department is Books and Sale is false
|
|
|
10
10
|
|
|
11
11
|
:evergreen_tree: Check out [IsoTree](https://github.com/ankane/isotree-ruby) for an alternative approach that uses Isolation Forest
|
|
12
12
|
|
|
13
|
-
[](https://github.com/ankane/outliertree-ruby/actions)
|
|
14
14
|
|
|
15
15
|
## Installation
|
|
16
16
|
|
data/lib/outliertree/result.rb
CHANGED
|
@@ -80,7 +80,7 @@ module OutlierTree
|
|
|
80
80
|
column: column,
|
|
81
81
|
value: value,
|
|
82
82
|
conditions: conditions,
|
|
83
|
-
group_statistics: group_statistics
|
|
83
|
+
group_statistics: group_statistics
|
|
84
84
|
# leave out for simplicity
|
|
85
85
|
# score: score,
|
|
86
86
|
# tree_depth: model_outputs.outlier_depth_final[row],
|
|
@@ -128,7 +128,7 @@ module OutlierTree
|
|
|
128
128
|
column: cond_col,
|
|
129
129
|
comparison: colcond,
|
|
130
130
|
to: condval,
|
|
131
|
-
value: colval
|
|
131
|
+
value: colval
|
|
132
132
|
# leave out for simplicity
|
|
133
133
|
# decimals: coldecim
|
|
134
134
|
}
|
data/lib/outliertree/version.rb
CHANGED
data/lib/outliertree.rb
CHANGED
|
@@ -5,10 +5,10 @@ require "outliertree/ext"
|
|
|
5
5
|
require "etc"
|
|
6
6
|
|
|
7
7
|
# modules
|
|
8
|
-
|
|
9
|
-
|
|
10
|
-
|
|
11
|
-
|
|
8
|
+
require_relative "outliertree/dataset"
|
|
9
|
+
require_relative "outliertree/model"
|
|
10
|
+
require_relative "outliertree/result"
|
|
11
|
+
require_relative "outliertree/version"
|
|
12
12
|
|
|
13
13
|
module OutlierTree
|
|
14
14
|
def self.new(**options)
|
|
@@ -11,7 +11,7 @@
|
|
|
11
11
|
* arXiv preprint arXiv:2001.00636 (2020).
|
|
12
12
|
*
|
|
13
13
|
*
|
|
14
|
-
* Copyright 2020 David Cortes.
|
|
14
|
+
* Copyright 2020-2024 David Cortes.
|
|
15
15
|
*
|
|
16
16
|
* Written for C++11 standard and OpenMP 2.0 or later. Code is meant to be wrapped into scripting languages
|
|
17
17
|
* such as R or Python.
|
|
@@ -97,6 +97,10 @@
|
|
|
97
97
|
* Model parameter. Default is 2.67.
|
|
98
98
|
* - z_outlier (in)
|
|
99
99
|
* Model parameter. Default is 8.0. Must be greater than z_norm.
|
|
100
|
+
* - check_nonneg_outliers (in)
|
|
101
|
+
* Whether to add an extra check for possible outliers defined as having negative values while all
|
|
102
|
+
* the rest have positive values, regardless of how many standard deviations away they are.
|
|
103
|
+
* This is currently only done on the first cluster (no conditions on any variable).
|
|
100
104
|
*
|
|
101
105
|
* Returns:
|
|
102
106
|
* - Whether there were any outliers detected.
|
|
@@ -107,7 +111,8 @@ bool define_numerical_cluster(double *restrict x, size_t *restrict ix_arr, size_
|
|
|
107
111
|
size_t cluster_num, size_t tree_num, size_t tree_depth,
|
|
108
112
|
bool is_log_transf, double log_minval, bool is_exp_transf, double orig_mean, double orig_sd,
|
|
109
113
|
double left_tail, double right_tail, double *restrict orig_x,
|
|
110
|
-
double max_perc_outliers, double z_norm, double z_outlier
|
|
114
|
+
double max_perc_outliers, double z_norm, double z_outlier,
|
|
115
|
+
bool check_nonneg_outliers)
|
|
111
116
|
{
|
|
112
117
|
|
|
113
118
|
/* TODO: this function could try to determine if the distribution is multimodal, and if so,
|
|
@@ -120,6 +125,7 @@ bool define_numerical_cluster(double *restrict x, size_t *restrict ix_arr, size_
|
|
|
120
125
|
/* NAs and Inf should have already been removed, and outliers with fewer conditionals already discarded */
|
|
121
126
|
bool has_low_values = false;
|
|
122
127
|
bool has_high_values = false;
|
|
128
|
+
bool has_outlier_neg_values = false;
|
|
123
129
|
long double running_mean = 0;
|
|
124
130
|
long double running_ssq = 0;
|
|
125
131
|
long double mean_prev = 0;
|
|
@@ -127,14 +133,15 @@ bool define_numerical_cluster(double *restrict x, size_t *restrict ix_arr, size_
|
|
|
127
133
|
double mean;
|
|
128
134
|
double sd;
|
|
129
135
|
size_t cnt;
|
|
130
|
-
size_t
|
|
136
|
+
size_t tot = end - st + 1;
|
|
137
|
+
size_t tail_size = (size_t) calculate_max_outliers((long double)tot, max_perc_outliers);
|
|
131
138
|
size_t st_non_tail = st + tail_size;
|
|
132
139
|
size_t end_non_tail = end - tail_size;
|
|
133
140
|
size_t st_normals = 0;
|
|
134
141
|
size_t end_normals = 0;
|
|
135
142
|
double min_gap = z_outlier - z_norm;
|
|
136
143
|
|
|
137
|
-
double curr_gap, next_gap,
|
|
144
|
+
double curr_gap, next_gap, lim_by_orig;
|
|
138
145
|
|
|
139
146
|
/* Note: there is no good reason and no theory behind these numbers.
|
|
140
147
|
TODO: find a better way of setting this */
|
|
@@ -166,9 +173,7 @@ bool define_numerical_cluster(double *restrict x, size_t *restrict ix_arr, size_
|
|
|
166
173
|
if ((!isinf(left_tail) || !isinf(right_tail)) && !is_log_transf && !is_exp_transf) {
|
|
167
174
|
sd *= 0.5;
|
|
168
175
|
}
|
|
169
|
-
sd = std::fmax(sd,
|
|
170
|
-
while (std::numeric_limits<double>::epsilon() > sd*std::fmin(min_gap, z_norm))
|
|
171
|
-
sd *= 4;
|
|
176
|
+
sd = std::fmax(sd, std::numeric_limits<double>::epsilon() / std::fmin(min_gap, z_norm));
|
|
172
177
|
cluster.cluster_mean = mean;
|
|
173
178
|
cluster.cluster_sd = sd;
|
|
174
179
|
cnt = end - st + 1;
|
|
@@ -212,10 +217,8 @@ bool define_numerical_cluster(double *restrict x, size_t *restrict ix_arr, size_
|
|
|
212
217
|
cluster.display_lim_low = orig_x[ix_arr[row + 1]];
|
|
213
218
|
cluster.perc_above = (long double)(end - st_normals + 1) / (long double)(end - st + 1);
|
|
214
219
|
|
|
215
|
-
|
|
216
|
-
|
|
217
|
-
cluster.lower_lim -= eps;
|
|
218
|
-
eps *= 4;
|
|
220
|
+
if (cluster.display_lim_low <= cluster.lower_lim) {
|
|
221
|
+
cluster.lower_lim = std::nextafter(cluster.display_lim_low, -std::numeric_limits<double>::infinity());
|
|
219
222
|
}
|
|
220
223
|
break;
|
|
221
224
|
}
|
|
@@ -225,6 +228,7 @@ bool define_numerical_cluster(double *restrict x, size_t *restrict ix_arr, size_
|
|
|
225
228
|
if (st_normals == 0) {
|
|
226
229
|
has_low_values = false;
|
|
227
230
|
} else {
|
|
231
|
+
assign_low_outliers:
|
|
228
232
|
for (size_t row = st; row < st_normals; row++) {
|
|
229
233
|
|
|
230
234
|
/* assign outlier if it's a better cluster than previously assigned */
|
|
@@ -254,7 +258,23 @@ bool define_numerical_cluster(double *restrict x, size_t *restrict ix_arr, size_
|
|
|
254
258
|
}
|
|
255
259
|
}
|
|
256
260
|
}
|
|
257
|
-
|
|
261
|
+
/* special type of outliers not based on standard deviations */
|
|
262
|
+
if (check_nonneg_outliers && st_normals == 0 && tot >= 500 && orig_x[ix_arr[st]] < 0. && orig_x[ix_arr[end]] >= 2.) {
|
|
263
|
+
size_t max_neg_outliers = (tot < 10000)? 1 : ((tot < 100000)? 2 : 3);
|
|
264
|
+
if (orig_x[ix_arr[st + max_neg_outliers]] > 0.) {
|
|
265
|
+
size_t num_neg = 0;
|
|
266
|
+
for (size_t row = st; row < st + max_neg_outliers; row++) {
|
|
267
|
+
num_neg += orig_x[ix_arr[row]] < 0.;
|
|
268
|
+
}
|
|
269
|
+
st_normals = st + num_neg;
|
|
270
|
+
cluster.lower_lim = 0.;
|
|
271
|
+
cluster.display_lim_low = orig_x[ix_arr[st + st_normals]];
|
|
272
|
+
cluster.perc_above = (long double)(end - st_normals + 1) / (long double)(end - st + 1);
|
|
273
|
+
has_outlier_neg_values = true;
|
|
274
|
+
goto assign_low_outliers;
|
|
275
|
+
}
|
|
276
|
+
}
|
|
277
|
+
if (!has_low_values && !has_outlier_neg_values) {
|
|
258
278
|
cluster.perc_above = 1.0;
|
|
259
279
|
if (!is_log_transf && !is_exp_transf) {
|
|
260
280
|
|
|
@@ -271,10 +291,8 @@ bool define_numerical_cluster(double *restrict x, size_t *restrict ix_arr, size_
|
|
|
271
291
|
}
|
|
272
292
|
|
|
273
293
|
if (cluster.lower_lim > -HUGE_VAL) {
|
|
274
|
-
|
|
275
|
-
|
|
276
|
-
cluster.lower_lim -= eps;
|
|
277
|
-
eps *= 4.;
|
|
294
|
+
if (cluster.lower_lim >= orig_x[ix_arr[st]]) {
|
|
295
|
+
cluster.lower_lim = std::nextafter(orig_x[ix_arr[st]], -std::numeric_limits<double>::infinity());
|
|
278
296
|
}
|
|
279
297
|
}
|
|
280
298
|
|
|
@@ -324,10 +342,8 @@ bool define_numerical_cluster(double *restrict x, size_t *restrict ix_arr, size_
|
|
|
324
342
|
cluster.display_lim_high = orig_x[ix_arr[row - 1]];
|
|
325
343
|
cluster.perc_below = (long double)(end_normals - st + 1) / (long double)(end - st + 1);
|
|
326
344
|
|
|
327
|
-
|
|
328
|
-
|
|
329
|
-
cluster.upper_lim += eps;
|
|
330
|
-
eps *= 4;
|
|
345
|
+
if (cluster.display_lim_high >= cluster.upper_lim) {
|
|
346
|
+
cluster.upper_lim = std::nextafter(cluster.display_lim_high, -std::numeric_limits<double>::infinity());
|
|
331
347
|
}
|
|
332
348
|
break;
|
|
333
349
|
}
|
|
@@ -384,10 +400,8 @@ bool define_numerical_cluster(double *restrict x, size_t *restrict ix_arr, size_
|
|
|
384
400
|
}
|
|
385
401
|
|
|
386
402
|
if (cluster.upper_lim < HUGE_VAL) {
|
|
387
|
-
|
|
388
|
-
|
|
389
|
-
cluster.upper_lim += eps;
|
|
390
|
-
eps *= 4.;
|
|
403
|
+
if (cluster.upper_lim <= orig_x[ix_arr[end]]) {
|
|
404
|
+
cluster.upper_lim = std::nextafter(orig_x[ix_arr[end]], std::numeric_limits<double>::infinity());
|
|
391
405
|
}
|
|
392
406
|
}
|
|
393
407
|
|
|
@@ -406,8 +420,8 @@ bool define_numerical_cluster(double *restrict x, size_t *restrict ix_arr, size_
|
|
|
406
420
|
}
|
|
407
421
|
|
|
408
422
|
/* save displayed statistics for cluster */
|
|
409
|
-
if (has_high_values || has_low_values || is_log_transf || is_exp_transf) {
|
|
410
|
-
size_t st_disp = has_low_values? st_normals : st;
|
|
423
|
+
if (has_high_values || has_low_values || is_log_transf || is_exp_transf || has_outlier_neg_values) {
|
|
424
|
+
size_t st_disp = (has_low_values || has_outlier_neg_values)? st_normals : st;
|
|
411
425
|
size_t end_disp = has_high_values? end_normals : end;
|
|
412
426
|
running_mean = 0;
|
|
413
427
|
running_ssq = 0;
|
|
@@ -428,7 +442,7 @@ bool define_numerical_cluster(double *restrict x, size_t *restrict ix_arr, size_
|
|
|
428
442
|
}
|
|
429
443
|
|
|
430
444
|
/* report whether outliers were found or not */
|
|
431
|
-
return has_low_values || has_high_values;
|
|
445
|
+
return has_low_values || has_high_values || has_outlier_neg_values;
|
|
432
446
|
}
|
|
433
447
|
|
|
434
448
|
|
|
@@ -11,7 +11,7 @@
|
|
|
11
11
|
* arXiv preprint arXiv:2001.00636 (2020).
|
|
12
12
|
*
|
|
13
13
|
*
|
|
14
|
-
* Copyright 2020 David Cortes.
|
|
14
|
+
* Copyright 2020-2024 David Cortes.
|
|
15
15
|
*
|
|
16
16
|
* Written for C++11 standard and OpenMP 2.0 or later. Code is meant to be wrapped into scripting languages
|
|
17
17
|
* such as R or Python.
|
|
@@ -552,7 +552,8 @@ void process_numeric_col(std::vector<Cluster> &cluster_root,
|
|
|
552
552
|
workspace.log_transf, workspace.log_minval, workspace.exp_transf,
|
|
553
553
|
workspace.orig_mean, workspace.orig_sd,
|
|
554
554
|
workspace.left_tail, workspace.right_tail, workspace.orig_target_col,
|
|
555
|
-
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier
|
|
555
|
+
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier,
|
|
556
|
+
true);
|
|
556
557
|
workspace.tree->back().clusters.push_back(0);
|
|
557
558
|
|
|
558
559
|
/* remove outliers if any were found */
|
|
@@ -636,7 +637,8 @@ void recursive_split_numeric(Workspace &workspace,
|
|
|
636
637
|
workspace.log_transf, workspace.log_minval, workspace.exp_transf,
|
|
637
638
|
workspace.orig_mean, workspace.orig_sd,
|
|
638
639
|
workspace.left_tail, workspace.right_tail, workspace.orig_target_col,
|
|
639
|
-
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier
|
|
640
|
+
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier,
|
|
641
|
+
false);
|
|
640
642
|
workspace.lev_has_outliers = workspace.has_outliers? true : workspace.lev_has_outliers;
|
|
641
643
|
|
|
642
644
|
if (model_params.follow_all && ((curr_depth + 1) < model_params.max_depth)) {
|
|
@@ -663,7 +665,8 @@ void recursive_split_numeric(Workspace &workspace,
|
|
|
663
665
|
workspace.log_transf, workspace.log_minval, workspace.exp_transf,
|
|
664
666
|
workspace.orig_mean, workspace.orig_sd,
|
|
665
667
|
workspace.left_tail, workspace.right_tail, workspace.orig_target_col,
|
|
666
|
-
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier
|
|
668
|
+
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier,
|
|
669
|
+
false);
|
|
667
670
|
workspace.lev_has_outliers = workspace.has_outliers? true : workspace.lev_has_outliers;
|
|
668
671
|
|
|
669
672
|
if (model_params.follow_all && ((curr_depth + 1) < model_params.max_depth)) {
|
|
@@ -687,7 +690,8 @@ void recursive_split_numeric(Workspace &workspace,
|
|
|
687
690
|
workspace.log_transf, workspace.log_minval, workspace.exp_transf,
|
|
688
691
|
workspace.orig_mean, workspace.orig_sd,
|
|
689
692
|
workspace.left_tail, workspace.right_tail, workspace.orig_target_col,
|
|
690
|
-
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier
|
|
693
|
+
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier,
|
|
694
|
+
false);
|
|
691
695
|
workspace.lev_has_outliers = workspace.has_outliers? true : workspace.lev_has_outliers;
|
|
692
696
|
|
|
693
697
|
if (model_params.follow_all && ((curr_depth + 1) < model_params.max_depth)) {
|
|
@@ -750,7 +754,8 @@ void recursive_split_numeric(Workspace &workspace,
|
|
|
750
754
|
workspace.log_transf, workspace.log_minval, workspace.exp_transf,
|
|
751
755
|
workspace.orig_mean, workspace.orig_sd,
|
|
752
756
|
workspace.left_tail, workspace.right_tail, workspace.orig_target_col,
|
|
753
|
-
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier
|
|
757
|
+
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier,
|
|
758
|
+
false);
|
|
754
759
|
workspace.lev_has_outliers = workspace.has_outliers? true : workspace.lev_has_outliers;
|
|
755
760
|
|
|
756
761
|
if (model_params.follow_all && ((curr_depth + 1) < model_params.max_depth)) {
|
|
@@ -777,7 +782,8 @@ void recursive_split_numeric(Workspace &workspace,
|
|
|
777
782
|
workspace.log_transf, workspace.log_minval, workspace.exp_transf,
|
|
778
783
|
workspace.orig_mean, workspace.orig_sd,
|
|
779
784
|
workspace.left_tail, workspace.right_tail, workspace.orig_target_col,
|
|
780
|
-
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier
|
|
785
|
+
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier,
|
|
786
|
+
false);
|
|
781
787
|
workspace.lev_has_outliers = workspace.has_outliers? true : workspace.lev_has_outliers;
|
|
782
788
|
|
|
783
789
|
if (model_params.follow_all && ((curr_depth + 1) < model_params.max_depth)) {
|
|
@@ -805,7 +811,8 @@ void recursive_split_numeric(Workspace &workspace,
|
|
|
805
811
|
workspace.log_transf, workspace.log_minval, workspace.exp_transf,
|
|
806
812
|
workspace.orig_mean, workspace.orig_sd,
|
|
807
813
|
workspace.left_tail, workspace.right_tail, workspace.orig_target_col,
|
|
808
|
-
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier
|
|
814
|
+
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier,
|
|
815
|
+
false);
|
|
809
816
|
workspace.lev_has_outliers = workspace.has_outliers? true : workspace.lev_has_outliers;
|
|
810
817
|
|
|
811
818
|
if (model_params.follow_all && ((curr_depth + 1) < model_params.max_depth)) {
|
|
@@ -871,7 +878,8 @@ void recursive_split_numeric(Workspace &workspace,
|
|
|
871
878
|
workspace.log_transf, workspace.log_minval, workspace.exp_transf,
|
|
872
879
|
workspace.orig_mean, workspace.orig_sd,
|
|
873
880
|
workspace.left_tail, workspace.right_tail, workspace.orig_target_col,
|
|
874
|
-
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier
|
|
881
|
+
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier,
|
|
882
|
+
false);
|
|
875
883
|
workspace.lev_has_outliers = workspace.has_outliers? true : workspace.lev_has_outliers;
|
|
876
884
|
|
|
877
885
|
if (model_params.follow_all && ((curr_depth + 1) < model_params.max_depth)) {
|
|
@@ -898,7 +906,8 @@ void recursive_split_numeric(Workspace &workspace,
|
|
|
898
906
|
workspace.log_transf, workspace.log_minval, workspace.exp_transf,
|
|
899
907
|
workspace.orig_mean, workspace.orig_sd,
|
|
900
908
|
workspace.left_tail, workspace.right_tail, workspace.orig_target_col,
|
|
901
|
-
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier
|
|
909
|
+
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier,
|
|
910
|
+
false);
|
|
902
911
|
workspace.lev_has_outliers = workspace.has_outliers? true : workspace.lev_has_outliers;
|
|
903
912
|
|
|
904
913
|
if (model_params.follow_all && ((curr_depth + 1) < model_params.max_depth)) {
|
|
@@ -928,7 +937,8 @@ void recursive_split_numeric(Workspace &workspace,
|
|
|
928
937
|
workspace.log_transf, workspace.log_minval, workspace.exp_transf,
|
|
929
938
|
workspace.orig_mean, workspace.orig_sd,
|
|
930
939
|
workspace.left_tail, workspace.right_tail, workspace.orig_target_col,
|
|
931
|
-
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier
|
|
940
|
+
model_params.max_perc_outliers, model_params.z_norm, model_params.z_outlier,
|
|
941
|
+
false);
|
|
932
942
|
workspace.lev_has_outliers = workspace.has_outliers? true : workspace.lev_has_outliers;
|
|
933
943
|
|
|
934
944
|
if (model_params.follow_all && ((curr_depth + 1) < model_params.max_depth)) {
|
|
@@ -733,7 +733,8 @@ bool define_numerical_cluster(double *restrict x, size_t *restrict ix_arr, size_
|
|
|
733
733
|
size_t *restrict outlier_depth, Cluster &cluster, std::vector<Cluster> &clusters, size_t cluster_num, size_t tree_num, size_t tree_depth,
|
|
734
734
|
bool is_log_transf, double log_minval, bool is_exp_transf, double orig_mean, double orig_sd,
|
|
735
735
|
double left_tail, double right_tail, double *restrict orig_x,
|
|
736
|
-
double max_perc_outliers, double z_norm, double z_outlier
|
|
736
|
+
double max_perc_outliers, double z_norm, double z_outlier,
|
|
737
|
+
bool check_nonneg_outliers);
|
|
737
738
|
void define_categ_cluster_no_cond(int *restrict x, size_t *restrict ix_arr, size_t st, size_t end, size_t ncateg,
|
|
738
739
|
double *restrict outlier_scores, size_t *restrict outlier_clusters, size_t *restrict outlier_trees,
|
|
739
740
|
size_t *restrict outlier_depth, Cluster &cluster,
|
metadata
CHANGED
|
@@ -1,14 +1,13 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: outliertree
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 0.
|
|
4
|
+
version: 0.4.1
|
|
5
5
|
platform: ruby
|
|
6
6
|
authors:
|
|
7
7
|
- Andrew Kane
|
|
8
|
-
autorequire:
|
|
9
8
|
bindir: bin
|
|
10
9
|
cert_chain: []
|
|
11
|
-
date:
|
|
10
|
+
date: 1980-01-02 00:00:00.000000000 Z
|
|
12
11
|
dependencies:
|
|
13
12
|
- !ruby/object:Gem::Dependency
|
|
14
13
|
name: rice
|
|
@@ -16,15 +15,14 @@ dependencies:
|
|
|
16
15
|
requirements:
|
|
17
16
|
- - ">="
|
|
18
17
|
- !ruby/object:Gem::Version
|
|
19
|
-
version: 4.
|
|
18
|
+
version: 4.3.3
|
|
20
19
|
type: :runtime
|
|
21
20
|
prerelease: false
|
|
22
21
|
version_requirements: !ruby/object:Gem::Requirement
|
|
23
22
|
requirements:
|
|
24
23
|
- - ">="
|
|
25
24
|
- !ruby/object:Gem::Version
|
|
26
|
-
version: 4.
|
|
27
|
-
description:
|
|
25
|
+
version: 4.3.3
|
|
28
26
|
email: andrew@ankane.org
|
|
29
27
|
executables: []
|
|
30
28
|
extensions:
|
|
@@ -60,7 +58,6 @@ homepage: https://github.com/ankane/outliertree-ruby
|
|
|
60
58
|
licenses:
|
|
61
59
|
- GPL-3.0-or-later
|
|
62
60
|
metadata: {}
|
|
63
|
-
post_install_message:
|
|
64
61
|
rdoc_options: []
|
|
65
62
|
require_paths:
|
|
66
63
|
- lib
|
|
@@ -68,15 +65,14 @@ required_ruby_version: !ruby/object:Gem::Requirement
|
|
|
68
65
|
requirements:
|
|
69
66
|
- - ">="
|
|
70
67
|
- !ruby/object:Gem::Version
|
|
71
|
-
version: '
|
|
68
|
+
version: '3.1'
|
|
72
69
|
required_rubygems_version: !ruby/object:Gem::Requirement
|
|
73
70
|
requirements:
|
|
74
71
|
- - ">="
|
|
75
72
|
- !ruby/object:Gem::Version
|
|
76
73
|
version: '0'
|
|
77
74
|
requirements: []
|
|
78
|
-
rubygems_version: 3.
|
|
79
|
-
signing_key:
|
|
75
|
+
rubygems_version: 3.6.7
|
|
80
76
|
specification_version: 4
|
|
81
77
|
summary: Explainable outlier/anomaly detection for Ruby
|
|
82
78
|
test_files: []
|