Question
Question Posted 02/14/20:
I am trying to create a machine learning data set from ADNI variables. The difficulty is knowing which variables are numeric and which are nominal. For example, PTEDUCAT seems like it would be nominal, but in PTDEMOG.csv, the column values are integers. Is there a table that identifies the data type of each variable. Ideally it would indicate whether the variable is meant to be numeric or nominal. Also, it would be helpful to know which variables are not fit for a machine learning data set. Such variables would be RID, VISCODE, PHASE, and any variables that have text as whole sentences.
I am trying to create a machine learning data set from ADNI variables. The difficulty is knowing which variables are numeric and which are nominal. For example, PTEDUCAT seems like it would be nominal, but in PTDEMOG.csv, the column values are integers. Is there a table that identifies the data type of each variable. Ideally it would indicate whether the variable is meant to be numeric or nominal. Also, it would be helpful to know which variables are not fit for a machine learning data set. Such variables would be RID, VISCODE, PHASE, and any variables that have text as whole sentences.
Response posted 02/18/20 by KC:
The data dictionary provided information about variable type.
