search

neurostuff / PyMARE

PyMARE: Python Meta-Analysis & Regression Engine
https://pymare.readthedocs.io
MIT License
51 stars 13 forks source link

Missing data imputation method , and small study effect #55

Closed safoora174 closed 3 years ago

safoora174 commented 3 years ago

Hi Is it possible to run missing data imputation method and small study effect in the package?

tsalo commented 3 years ago

@safoora174 to clarify, do you want these methods for neuroimaging meta-analysis, or for meta-analysis in general?

safoora174 commented 3 years ago

@tsalo For meta-analysis in general

tsalo commented 3 years ago

Ah, thank you. In that case, I believe this issue is better suited for the PyMARE repository. NiMARE is for neuroimaging meta-analysis, while PyMARE is for standard meta-analysis methods. I will transfer the issue.

safoora174 commented 3 years ago

@tsalo thank you, I thin I post it in a wrong place

tyarkoni commented 3 years ago

@safoora174 there aren't currently any plans to support imputation in PyMARE; it seems out of scope. There are other Python packages you can use for imputation—e.g., see the scikit-learn docs. Once you have an imputed dataset, you can then feed it to PyMARE.

safoora174 commented 3 years ago

@tyarkoni thank you so much. I have read nearly all the missing data imputation packages in python, such as autoimpute, missingpy,impute , but non of them using methods which are used in metamiss function of metasense package in R. They do not use methods such as ACA, IMOR, GH. Thanks for your comment

safoora174 commented 3 years ago

Dear Taylor,

I wrote some of missing data Imputation method myself and after analysis, I have the summary risk ratios for each analyses. Do you know how I can combine them in one forest plot? I searched a lot but couldn't find a useful method. I found arviz , but I do not know how combine them in one plot and how I can change the code to change the shape of each study plot to a diamond one. Sorry for the confusion

Regards, Safoora

On Wed, Nov 25, 2020, 6:27 PM Tal Yarkoni notifications@github.com wrote:

Closed #55 https://github.com/neurostuff/PyMARE/issues/55.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/neurostuff/PyMARE/issues/55#event-4038492954, or unsubscribe https://github.com/notifications/unsubscribe-auth/APWN25XZA7IRRIJELB5MFRTSRULMRANCNFSM4T6CIQNA .

tsalo commented 3 years ago

Hi Safoora,

I'm sorry to say that standard meta-analyses are mostly outside of my area of expertise. I don't know how to generate forest plots for summary risk ratios, or even generally, in Python.

safoora174 commented 3 years ago

Thank you so much dear Taylor.

On Wed, Dec 9, 2020 at 11:31 PM Taylor Salo notifications@github.com wrote:

Hi Safoora,

I'm sorry to say that standard meta-analyses are mostly outside of my area of expertise. I don't know how to generate forest plots for summary risk ratios, or even generally, in Python.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/neurostuff/PyMARE/issues/55#issuecomment-742014984, or unsubscribe https://github.com/notifications/unsubscribe-auth/APWN25UI45YCFCIUHDZEASTST7JTHANCNFSM4T6CIQNA .