Error in if (nrow(d) == 0) return(NULL) : argument is of length zero

[ikram2779]

extract_instruments("ukb-b-7992", access_token = NULL)
Error in if (nrow(d) == 0) return(NULL) : argument is of length zero

please help

[PPKstar]

I meet the problem too. And when I use "extract_outcome_data", it just return NULL.

[PPKstar]

I have tried to update "TwoSampleMR" package, but it didn't work.

[zhengjinjinjin]

Me too!

[Ali-mhm]

I meet the problem too. And when I use "extract_outcome_data", it just return NULL.

in the TWOsampleMR package whenever you receive a weird error, it is probably a server issue such as this error, or when you are doing clump, that uses this line of code:

options(ieugwasr_api = 'gwas-api.mrcieu.ac.uk/')

[Gseven222]

I meet the problem too.

[DC-Jer]

Confront with same problem from yesterday afternoon when using the API options to query the data.
(https://github.com/MRCIEU/TwoSampleMR/assets/37207740/133fc0fb-4cb4-4913-8685-e10d7ca9586b)

[fcf01]

I meet this issue too, and still undo now, anybody could help me
expo_rt<-extract_instruments(outcome= "balabalabla,p1 = 1e-5,

•                                                        clump = T,
  

•                                                        p2 = 1e-5,
  

•                                                        r2 = 0.001,
  

•                                                        kb = 10000) 
  

and tell me : Error in if (nrow(d) == 0) return(NULL) : argument is of length zero

[zhuhhfsdfsd]

I meet the problem too.

[wuyisky]

I meet the problem too.

[cykky]

something wrong with the server?

[Spasmodic114]

Yesterday I accessed the website (https://gwas.mrcieu.ac.uk) and it showed Server Error 500. Now I can access the website but I met the same problem with TwoSampleMR package. :(

[xiaobaiyaobianqiang]

I meet the problem too.

[Castiel4517]

Me too, is that a bug?

[xiaobaiyaobianqiang]

how to solve

[ikram2779]

it is working today as usual. probably it was server issue

[Spasmodic114]

I met the problem again 1 hour ago :(
I wonder if it is the server error or other problems?

[Spasmodic114]

It showed "server code: 502" before: "Error in if (nrow(d) == 0) return(NULL) : argument is of length zero"

[xiaobaiyaobianqiang]

I meet the problem too.how to solve

[crebli]

me too even I got the access token.

[xiaobaiyaobianqiang]

I solve the question by
options(ieugwasr_api = 'gwas-api.mrcieu.ac.uk/')

[cmulph]

Thank U

[Fin-beep]

Hello, what are the specific steps

[Lecter-HHH]

Thank you! This method works!

[SUMCZY2021]

this method not work for me. i still get "Error in if (nrow(d) == 0) return(NULL) : argument is of length zero". does anyone have other solution?

[SUMCZY2021]

che<-extract_instruments(outcomes='ukb-b-1489', clump=TRUE)
Error in if (nrow(d) == 0) return(NULL) : argument is of length zero".

[crebli]

Nice! Thank you!发自我的荣耀手机-------- 原始邮件 --------发件人： xiaobaiyaobianqiang ***@***.***>日期： 2024年5月5日周日 下午4:52收件人： MRCIEU/TwoSampleMR ***@***.***>抄送： crebli ***@***.***>, Comment ***@***.***>主 题： Re: [MRCIEU/TwoSampleMR] Error in if (nrow(d) == 0) return(NULL) : argument is of length zero (Discussion #491) I solve the question by options(ieugwasr_api = 'gwas-api.mrcieu.ac.uk/') —Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you commented.Message ID: ***@***.***>

[vividchan]

I also solved this question by your method, thanks

[yuwei0529]

thank you very much,i solved too

[Fin-beep]

Hello, what are the specific steps

[a1141539455]

I met this problem even I got the token：
options(ieugwasr_api = 'gwas-api.mrcieu.ac.uk/')

[vividchan]

You can try to use another dataset to validate whether your previous dataset is not readable. Because I find out that some datasets are not available.

[a1141539455]

I have tried. I was able to extract the instrumental variable data set before, but now I keep getting this error. :(
Can you use it normally now?

[vividchan]

Yes, the code works under my R enviroment. Maybe you can try to re-download the ieugwasr package.

[a1141539455]

I have re-installed ieugwasr1.0.0 and twosamplemr0.6.2 but it still gives the error :(

[ManqiZh]

The same problem is repeated today.

[YzGLab]

how to slove it , it repeat all days!

[AliiReezaa]

options(ieugwasr_api = 'gwas-api.mrcieu.ac.uk/') This method was working for me for a long time; unfortunately, starting yesterday, it is no longer working. Do you have any idea how to resolve this issue?

[remlapmot]

We have told users many times to please stop using the old API address - there is no need to set this in options() anymore.

[AliiReezaa]

So how to do it if we are not supposed to use this API address? What is the new one?

[remlapmot]

With the latest versions of TwoSampleMR and ieugwasr it is already set for you - just follow the authentication steps to setup a token - https://mrcieu.github.io/ieugwasr/articles/guide.html#authentication .

[Ali-mhm]

With the latest versions of TwoSampleMR and ieugwasr it is already set for you - just follow the authentication steps to setup a token - https://mrcieu.github.io/ieugwasr/articles/guide.html#authentication .

You are right but
For two weeks, whenever you try to obtain a token and sign in, it results in an overload error.

[ldbldb]

我们已经多次告诉用户，请停止使用旧的 API 地址 - 无需再设置它。options()

使用最新版本的 TwoSampleMR 和 ieugwasr，它已经为您设置好了 - 只需按照身份验证步骤设置令牌 - https://mrcieu.github.io/ieugwasr/articles/guide.html#authentication 。

Hello, using the latest version of TwoSampleMR and ieugwasr, I still get the error: error in if (out == “[]”) {: parameter length is zero, then I can get the result after changing the parameter to: clump = FALSE, may I ask what is the reason?

[AliiReezaa]

I appreciate your informative guide. It finally worked. best regards, Ali Reza

…

On Mon, Jan 13, 2025 at 5:23 PM Ali-mhm ***@***.***> wrote: With the latest versions of TwoSampleMR and ieugwasr it is already set for you - just follow the authentication steps to setup a token - https://mrcieu.github.io/ieugwasr/articles/guide.html#authentication . You are right but For two weeks, whenever you try to obtain a token and sign in, it results in an overload error. — Reply to this email directly, view it on GitHub <#491 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BCVDRBO3KGNGPVMSTSL5QIL2KOAXPAVCNFSM6AAAAABFEWY326VHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTCOBRG44DQNQ> . You are receiving this because you commented.Message ID: ***@***.***>

[fe1m22]

I have been having difficulties connecting to the API using extract_instruments as part of TwoSampleMR package. I have tried new tokens, re-installing the ieugwasr and TwoSampleMR packages and updating R studio. Any advice would be very helpful. I get the following error:

Error in if (out == "[]") { : argument is of length zero

[fe1m22]

I have just read the above comment and clump=F seems to work but then the SNPs are not clumped. Does anyone know a work around? Any help would be appreciated.

[ldbldb]

After modifying clump=F, in the next step just clump the exposed SNPs and merge them with the outcome SNPs , but I'm not sure if this is the right way to go about it

[HuXuantao]

I'm pretty sure there is sth wrong with the server or the package code. I tried to debug for a whole day and here is the findings:

1. When using default args, you can get the normal returns from functions like extract_outcome etc. (I compared a new returns with my previous one produced by a successful run)
2. BUT if you change the default setup of args (for me, i tried to merely change r2 = 0.001 to r2 = 0.0011), you will get a "(nrow(d) == 0) return(NULL)". (I also compared a run using the same args with one of my previous successful run. it just failed.)

I have already updated twosamplemr (0.6.11) and ieugwasr (1.0.1) package and got my IEU token working. But these seems having nothing to do with the ERROR. lol!
Frustrating...

[HuXuantao]

Today, it works...