search

Tirgit / missCompare

missCompare R package - intuitive missing data imputation framework
Other
39 stars 5 forks source link

help wanted #3

Closed sandeshregmi closed 3 years ago

sandeshregmi commented 5 years ago

is there a way to fix this issue? missCompare::MCAR(simulated$Simulated_matrix, MD_pattern = metadata$MD_Pattern, NA_fraction = metadata$Fraction_missingness, min_PDM = 10) 27.62% of observations (with at least one missing datapoint) covered by setting min_PDM to 10 Error: Proportion of missing cells is too large in combination with the desired number of missing variables

Tirgit commented 5 years ago

Hi - the only way to "fix" this is to try different values for the min_PDM argument. I suggest you run clean (with your sensible, pre-determined variable and individual thresholds) and then missCompare::get_data() on your dataset and check the outcome object called min_PDM_thresholds. Perhaps you will need to pick a lower number to retain more information. Please also study the vignette for the package for a detailed tutorial: vignette("misscompare") Let me know if these steps work for you!