alexchristensen / semnet Goto Github PK

Hi,

I am using the SemNeT package to calculate semantic networks of two datasets. In each datasets two groups are compared and the two files look the same, apart from the animals the participants said. The package works really great with one of the datasets, but not with the other one. For the one with problems I can estimate the networks, compare them to a random network, but I get an error for the bootstrap analysis. The function test.bootSemNeT leads to the error: "Warning: Error in contrasts<-: contrasts can be applied only to factors with 2 or more levels" However, the bootstrap graphs are plotted and they show two groups. This error also only appears if I use the TMFG method with Pearson's correlation. Is there anything in a data set that can lead to such a mistake?

Best wishes,
Michaela

Copy and paste of error from your R console

Computing forward flow with glove...
Computing forward flow with glove...
Computing forward flow with glove...
Computing forward flow with glove...
Computing forward flow with glove...
Computing forward flow with glove...
Error: node stack overflow
Error during wrapup: node stack overflow
Error: no more error handlers available (recursive errors?); invoking 'abort' restart

To Reproduce:

• Function error occurred in: ForwardFlow.R
  -#import data
  response_matrix <- read.csv("response_matrix.csv")
  #run function
  pilot_results <- forward_flow(
  response_matrix = response_matrix,
  semantic_space = "all",
  min_cue = min_cue,
  min_response = min_response,
  max_response = max_response,
  type = "free")

Optional (but extremely helpful): Attach data to issue or send data to [email protected] with the subject line SemNeT Data: [your GitHub username] using this template
response_matrix.csv

R and SemNeT versions:

• R: 4.2.1
• SemNeT: 1.4.4

Operating System:

• OS: Mac
• Version 12.2.1

Additional context and comments

• I also tried running the function limited the data to 1 participant and 1 semantic space but received the same error.

Copy and paste of error from your R console

Warning: Navigation containers expect a collection of `bslib::nav()`/`shiny::tabPanel()`s and/or `bslib::nav_menu()`/`shiny::navbarMenu()`s. Consider using `header` or `footer` if you wish to place content above (or below) every panel's contents.
Warning in spreadr::spreadr(network = nets[[net_name]], start_run = act_df,  :
  These nodes specified in start_run don't exist in network: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112

To Reproduce:

• Function error occurred in: [e.g., SemNeT Shiny Spreading Activation]

Optional (but extremely helpful): Attach data to issue or send data to [email protected] with the subject line SemNeT Data: [your GitHub username] using this template

R and SemNeT versions:

• R: [e.g., 3.6.3]
• SemNeT: [e.g., 1.3.0]

Operating System:

• OS: [Windows]
• Version [10]

Additional context and comments

• [In the load data tab of the shiny app it asks for the BRM but when I upload that data file the semantic network analysis doesn't work. In order for me to get the network analysis to work, I need to uploaded the cleaned animal responses. I am not sure if this is what causes the node error later on in the SpreadR analysis]
• [The Spreading Activation Analysis seems to work when looking at the visualization but the result output is 0 for all nodes across all timepoints which leads me to believe that there is an error. Based on Siew et al., 2019, I set the decay = 0.5, retention and suppression = 0. As in the Siew paper we arbitrarily picked 20 as our activation at time=0. ]
• [For the semantic network portion, we used the TMFG filtering, cosine similarity, and minimum response of 2 participants]
• [I ran this spreadr with Cynthia Siew's r scripts available on her OSF site but based on Reviewer feedback we would like to run this spreading activation analysis successfully in SemNeT.]
  [HCMS_Animals_cleaned_responses.csv](https://github.com/AlexChristensen/SemNeT/files/11022874/HCM
  group.csv
  S_Animals_cleaned_responses.csv)
  HCMS_Animals_BRM.csv

Copy and paste of error from your R console

Error in compare.nets(Young.net, Old.net, title = list("Younger Adults",  : 
  could not find function "compare.nets"

To Reproduce:

• Function error occurred in: [SemNet]

Optional (but extremely helpful): Attach data to issue or send data to [email protected] with the subject line SemNeT Data: [your GitHub username] using this template

R and SemNeT versions:

• R: ["R version 4.0.3 (2020-10-10)"]
• SemNeT: [‘1.4.1’]

Operating System:

• OS: [Windows]
• Version [10 x64 build 18363]

Additional context and comments

• [# Visually compare networks
  compare.nets(Young.net, Old.net,
  title = list("Younger Adults", "Older Adults"),
  config = "spring", weighted = TRUE)]

• [worked before but doesn't now]

Copy and paste of error from your R console

SemNetShiny()

To Reproduce:

• Function error occurred in: SemNetShiny

R and SemNeT versions:

• R: 4.0.3
• SemNeT: 1.3.0

Don't know your versions? You can check using the following code:

# R
R.version$version.string
# SemNeT
packageVersion("SemNeT")

Operating System:

• OS: Mac
• Version Darwin 19.6.0
  freezeExample.zip

Don't know your OS? You can check using the following code:

# OS
Sys.info()[1:3]

Additional context and comments

• [e.g., function with arguments]
• [e.g., worked before but doesn't now]
• [e.g., expected behavior]
• [e.g., other things you've tried]

Warning: Error in [: subscript out of bounds

SemNeTShiny()

To Reproduce:

• Function error occurred in: [e.g., when I would like to estimate network of the method after attaching the textcleaner object variable]

Optional (but extremely helpful): Attach data to issue or send data to [email protected] with the subject line SemNeT Data: [your GitHub username] using this template

R and SemNeT versions:

• R:4.1.3
• SemNeT: [e.g., 1.4.3]

**Operating System:**
 - OS:  [Mac]
 - Version [10.15.7, 13-inch, 2018]

--->

**Additional context and
 comments**
[I put the text cleaner object variable into shiny, and then I try to estimate the network using 4 methods, but neither of them works]
[Attchment includes the data after cleaning
[cleandata.csv](https://github.com/AlexChristensen/SemNeT/files/8457872/cleandata.csv)
]

Copy and paste of error from your R console

install.packages('SemNeTShiny')
Installing package into ‘/Users/mtd143/Library/R/4.0/library’
(as ‘lib’ is unspecified)
Warning in install.packages :
  package ‘SemNeTShiny’ is not available for this version of R

To Reproduce:

• Function error occurred in: [R and RStudio after installing SemNeT 1.3.0 and shiny 1.6.0]

Optional (but extremely helpful): Attach data to issue or send data to [email protected] with the subject line SemNeT Data: [mtd143] using this template

R and SemNeT versions:

• R: [4.0.3]
• SemNeT: [e.g., 1.3.0]

Operating System:

• OS: [Mac]
• Version [Mojave 10.14.6 (18G8022)]

Additional context and comments

• [I tried installing in both R and R studio, but get the same error]

Hi Alex,
In your SemNA Tutorial paper with Yoed Kenett, you mention that you can use the general dictionary for phonological verbal fluency tasks. My question is, what do you use to create the edges for the phonological network? Siew (2013) and Vitevitch (2002,2008) used phonological similarity between words to determine edges. I did not see specifically what you would use to calculate the edges/edge wights from the SemNet package, but please let me know if I missed something.
Thank you,
Abby

Hi @AlexChristensen , do you plan on fixing the CRAN check issues and resurrect the package on CRAN? I'm asking because the package is currently still listed in the Psychometrics task view, see: cran-task-views/Psychometrics#26
Thanks in advance for your consideration!

Copy and paste of error from your R console
Error in stat.cooccur(data, window = window, alpha = alpha):
CN(): Only a response matrix or ordered numeric matrix can be used as input for 'data'
Copy and paste error code here
net.low <- CN(low)

To Reproduce:

• Function error occurred in: CN

Optional (but extremely helpful): Attach data to issue or send data to [email protected] with the subject line SemNeT Data: ezequielfk using this template

R and SemNeT versions:

• R: 4.1.0
• SemNeT: 1.4.3

Operating System:

• OS: Windows
• Version: 11

Additional context and comments

• I tried running the shiny app SemNeTShiny(). Only in the case of using the steps in the manuscript with the "open.animals" database have I been able to estimate the corresponding networks. However, I have problems when I want to do the analysis with a different database, or when I want to apply the steps directly in R instead of the shiny app.
  In the latter case, I follow the steps in supplementary material 6 (SI6), and I get the following error when running the function net.low <- CN(low):

net.low <- CN(low)
Error in stat.cooccur(data, window = window, alpha = alpha):
CN(): Only a response matrix or ordered numeric matrix can be used as input for 'data'

A similar problem occurs when I want to incorporate another binary database. In the case of the shiny app, the program stops when trying to load a binary data base (for example, a database of mine, or the "high" database derived from the open.animals database). When I want to incorporate or estimate the network of a binary database, the program crashes (or the corresponding function returns such an error).
My main purpose is to apply the network estimation methods you propose, but from a pre-existing binary database. That is my big obstacle, mainly that when I try to load a binary database to the shiny app the program crashes or gives me the error indicated in the previous message (if I directly run the syntax in R).
Well I await any help on how to implement the methods you propose but directly on a binary database like the ones I sent you attached (i.e. without the need to select from clean$response in the shiny app).
Thanks!
high_BRM.csv
low_BRM.csv
prueba.csv