Skip to content
This repository has been archived by the owner on Nov 13, 2021. It is now read-only.

AnomalyDetectionTs I got a blank plot window #94

Open
aliwumi opened this issue Dec 21, 2017 · 4 comments
Open

AnomalyDetectionTs I got a blank plot window #94

aliwumi opened this issue Dec 21, 2017 · 4 comments

Comments

@aliwumi
Copy link

aliwumi commented Dec 21, 2017

I'm a newbie in R, i got a problem when I am trying to detect anomalies on my test dataset use AnomalyDetection‘s function AnomalyDetectionTs.

My R code and the the plot window is blank:
problem

My test data:
time,rate
2017-08-01 00:00:00,0.7742
2017-08-02 00:00:00,0.7548
2017-08-03 00:00:00,0.7548
2017-08-04 00:00:00,0.7087
2017-08-05 00:00:00,0.7425
2017-08-06 00:00:00,0.7321
2017-08-07 00:00:00,0.7757
2017-08-08 00:00:00,0.7478
2017-08-09 00:00:00,0.745
2017-08-10 00:00:00,0.7264
2017-08-11 00:00:00,0.7139
2017-08-12 00:00:00,0.7257
2017-08-13 00:00:00,0.7307
2017-08-14 00:00:00,0.7443
2017-08-15 00:00:00,0.6525
2017-08-16 00:00:00,0.6732
2017-08-17 00:00:00,0.7324
2017-08-18 00:00:00,0.7288
2017-08-19 00:00:00,0.7485
2017-08-20 00:00:00,0.6779
2017-08-21 00:00:00,0.7564
2017-08-22 00:00:00,0.7322
2017-08-23 00:00:00,0.7337
2017-08-24 00:00:00,0.6972
2017-08-25 00:00:00,0.7336
2017-08-26 00:00:00,0.737
2017-08-27 00:00:00,0.716
2017-08-28 00:00:00,0.7535
2017-08-29 00:00:00,0.7557
2017-08-30 00:00:00,0.7318
2017-08-31 00:00:00,0.7754
2017-09-01 00:00:00,0.7403
2017-09-02 00:00:00,0.7304
2017-09-03 00:00:00,0.7211
2017-09-04 00:00:00,0.748
2017-09-05 00:00:00,0.7308
2017-09-06 00:00:00,0.7473
2017-09-07 00:00:00,0.7372
2017-09-08 00:00:00,0.7663
2017-09-09 00:00:00,0.7455
2017-09-10 00:00:00,0.75
2017-09-11 00:00:00,0.7514
2017-09-12 00:00:00,0.7194
2017-09-13 00:00:00,0.7211
2017-09-14 00:00:00,0.7388
2017-09-15 00:00:00,0.7595
2017-09-16 00:00:00,0.7438
2017-09-17 00:00:00,0.7495
2017-09-18 00:00:00,0.7681
2017-09-19 00:00:00,0.7409
2017-09-20 00:00:00,0.7249
2017-09-21 00:00:00,0.7355
2017-09-22 00:00:00,0.7611
2017-09-23 00:00:00,0.7672
2017-09-24 00:00:00,0.7474
2017-09-25 00:00:00,0.7472
2017-09-26 00:00:00,0.7324
2017-09-27 00:00:00,0.7957
2017-09-28 00:00:00,0.7586
2017-09-29 00:00:00,0.8076
2017-09-30 00:00:00,0.7578
2017-10-01 00:00:00,0.7403
2017-10-02 00:00:00,0.7722
2017-10-03 00:00:00,0.7634
2017-10-04 00:00:00,0.7516
2017-10-05 00:00:00,0.776
2017-10-06 00:00:00,0.7908
2017-10-07 00:00:00,0.7786
2017-10-08 00:00:00,0.7484
2017-10-09 00:00:00,0.7665
2017-10-10 00:00:00,0.7503
2017-10-11 00:00:00,0.7739
2017-10-12 00:00:00,0.7928
2017-10-13 00:00:00,0.7717
2017-10-14 00:00:00,0.8009
2017-10-15 00:00:00,0.7931
2017-10-16 00:00:00,0.7905
2017-10-17 00:00:00,0.7981
2017-10-18 00:00:00,0.7963
2017-10-19 00:00:00,0.7726
2017-10-20 00:00:00,0.8056
2017-10-21 00:00:00,0.8236
2017-10-22 00:00:00,0.7712
2017-10-23 00:00:00,0.7942
2017-10-24 00:00:00,0.7782
2017-10-25 00:00:00,0.8123
2017-10-26 00:00:00,0.7928
2017-10-27 00:00:00,0.7891
2017-10-28 00:00:00,0.8051
2017-10-29 00:00:00,0.7841
2017-10-30 00:00:00,0.7815
2017-10-31 00:00:00,0.8014
2017-11-01 00:00:00,0.8114
2017-11-02 00:00:00,0.7825
2017-11-03 00:00:00,0.7902
2017-11-04 00:00:00,0.8024
2017-11-05 00:00:00,0.8012
2017-11-06 00:00:00,0.8135
2017-11-07 00:00:00,0.781
2017-11-08 00:00:00,0.8088
2017-11-09 00:00:00,0.7858
2017-11-10 00:00:00,0.7437
2017-11-11 00:00:00,0.7529
2017-11-12 00:00:00,0.7734
2017-11-13 00:00:00,0.808
2017-11-14 00:00:00,0.7459
2017-11-15 00:00:00,0.692
2017-11-16 00:00:00,0.7114
2017-11-17 00:00:00,0.6866
2017-11-18 00:00:00,0.7112
2017-11-19 00:00:00,0.7282
2017-11-20 00:00:00,0.7214
2017-11-21 00:00:00,0.7705
2017-11-22 00:00:00,0.7705
2017-11-23 00:00:00,0.7844
2017-11-24 00:00:00,0.7603
2017-11-25 00:00:00,0.774
2017-11-26 00:00:00,0.7562
2017-11-27 00:00:00,0.7923
2017-11-28 00:00:00,0.7687
2017-11-29 00:00:00,0.7698
2017-11-30 00:00:00,0.7812
2017-12-01 00:00:00,0.765
2017-12-02 00:00:00,0.7837
2017-12-03 00:00:00,0.7863
2017-12-04 00:00:00,0.787
2017-12-05 00:00:00,0.7888
2017-12-06 00:00:00,0.7811
2017-12-07 00:00:00,0.7525
2017-12-08 00:00:00,0.817
2017-12-09 00:00:00,0.8083
2017-12-10 00:00:00,0.8361
2017-12-11 00:00:00,0.8086
2017-12-12 00:00:00,0.8022
2017-12-13 00:00:00,0.8041
2017-12-14 00:00:00,0.8016
2017-12-15 00:00:00,0.8131
2017-12-16 00:00:00,0.8331
2017-12-17 00:00:00,0.8265
2017-12-18 00:00:00,0.8253
2017-12-19 00:00:00,0.8175

@danielmedbar
Copy link

Hi there!
I get the same. It seems this happens when you have plot=T and 0 anomalies detected.
For some reason instead of showing the plot with 0 outliers it produces a blank plot.

Does anyone here know how to sort this out?

@IronistM
Copy link

IronistM commented Jan 23, 2018

+1. This happens when no output is supplied to the plot. Probably something like this would solve in in the _ts function.

if(plot & length(anoms) > 0){
   return (list(anoms = anoms, plot = xgraph))
 } else {
   return (list(anoms = anoms, plot = plot.new()))
 }

@danielmedbar
Copy link

i think the problem is there and in line 148:
if (anom_pct == 0) { if (verbose) message("No anomalies detected.") return(list(anoms = data.frame(), plot = plot.new())) }
it should have to be changed to:
if (anom_pct == 0) { if (verbose) message("No anomalies detected.") return(list(anoms = data.frame())) }

@HanWang-HiWorld
Copy link

HanWang-HiWorld commented Apr 18, 2018

AnomalyDetectionTs is very sensitive to max_anoms and longterm setting. This is how I fixed the issue with more than one month time series:
anomalydetection <- AnomalyDetectionTs(data, max_anoms=0.02, direction = "pos", plot = TRUE, e_value = T, longterm = T)

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants