Closed bpoblador closed 1 year ago
Hi @bpoblador, this is very strange! Did this happen after entering the new meaning of visits? Has something changed in the INSTANCE table?
could you please check the file D3_Stream_PROMPTS.html in the folder _g_describeHTML/02 and D3_Stream_PROMPTS_check.html in the folder _g_describeHTML/03? these files provide you with a description of the records from the PROMPTs before and after the consistency checks
Giorgio
Hi @GiorgioLimoncella ,
Yes, this happend after the modifications the other times I have run this algorithm the prompts has been included.
D3_Stream_PROMPTS.html doesn't exist in the folder g_describe_HTML/02 and neither does D3_Stream_PROMPTS_check.html in the folder g_describe_HTML/03.
In both folders appear the files regarding conceptsets and itemsets, but the last one doesn't appear in the graph.
Bea
@bpoblador could you please check if also SURVEY_ID_BR.html is missing? (in g_describe_HTML/01_01)
many thanks!
@GiorgioLimoncella It is not missing, it is there:
@bpoblador
if you open it you should find a description of how many records were detected from the _surveiid, could you please check and tell me if the numbers seem correct to you? (approximately)
if the numbers look correct the problem should be in _step_02_01_T2.2_create_pregnancies_fromprompts.R, I'm checking it now!
Yes, the numbers seem correct. We are going to check if any error is produced when we run the file step_02_01_T2.2_create_pregnancies_from_prompts.R,
After checking and running step_02_01_T2.2_create_pregnancies_from_prompts.R, it looks like the issue for catching the promt are in the merge of line 42, adding to the merge the parameter allow.cartesian = T seems to solve the problem. Now, it creates D3_Stream_PROMPTS.html in in the folder g_describe_HTML/02. Most of the data shown in the html look like fine, but there are some parts with a lot of missing data in the report. I am still checking if everything is solved
Hi @asanmej, @bpoblador,
this is strange. The merge is done using the survey_id as the key and there seems to be a lot of repeated survey_ids. If you want we can arrange a call and check the problem together!
Hi @GiorgioLimoncella, if you have availability this week we can meet and see what is happening. I'm sending our avaliability by email.
You were right, we had duplicated records. Once fixed it runs without errors.
Thanks!
Dear @GiorgioLimoncella and @clobart
After run the algorithm, the results show that only conceptset data has been used, but we have also prompts. Do you know why this happend?
Bea