emzed.stats.oneWayAnova(factorColumn, dependentColumn)[source]
emzed.stats.kruskalWallis(factorColumn, dependentColumn)[source]

see Statistical Analysis for example usage

emzed.stats.oneWayAnovaOnTables(tableSet1, tableSet2, idColumn, valueColumn)[source]

Compares two sets of tables. Each set is a list of tables, with common columns idColumn and valueColumn. The first one is a factor which used to build groups, the latter is the dependent numerical value.

Eg you have to lists with tables, where each table has factor column compound and dependent value column foldChange. Then you get a result table which looks like:

>>> tresult = emzed.stats.oneWayAnovaOnTables(tables1, tables2, idColumn="compound", valueColumn="foldChange") 
>>> print tresult
id       n1       n2       avg1_foldChange std1_foldChange avg2_foldChange std2_foldChange p_value 
str      int      int      float           float           float           float           float   
------   ------   ------   ------          ------          ------          ------          ------  
ATP      4        6        1.40            0.40            0.40            0.30            0.90    
ADP      5        6        1.60            0.13            1.50            0.08            0.23    

emzed.stats.kruskalWallisOnTables(tableSet1, tableSet2, idColumn, valueColumn)[source]

Works as oneWayAnovaOnTables() above, but uses non parametric kruskal wallis test.