I am trying to apply TukeyHSD from statsmodels but receive the following error message.
---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-68-503f7cac5cbb> in <module>()
1 from statsmodels.stats.multicomp import pairwise_tukeyhsd
2
----> 3 pairwise_tukeyhsd(endog=data['value'], groups=data['sample_id'])
c:\users\hambarsous\appdata\local\programs\python\python36\lib\site-packages\statsmodels\stats\multicomp.py in pairwise_tukeyhsd(endog, groups, alpha)
36 '''
37
---> 38 return MultiComparison(endog, groups).tukeyhsd(alpha=alpha)
c:\users\hambarsous\appdata\local\programs\python\python36\lib\site-packages\statsmodels\sandbox\stats\multicomp.py in __init__(self, data, groups, group_order)
794 if group_order is None:
795 self.groupsunique, self.groupintlab = np.unique(groups,
--> 796 return_inverse=True)
797 else:
798 #check if group_order has any names not in groups
c:\users\hambarsous\appdata\local\programs\python\python36\lib\site-packages\numpy\lib\arraysetops.py in unique(ar, return_index, return_inverse, return_counts, axis)
221 ar = np.asanyarray(ar)
222 if axis is None:
--> 223 return _unique1d(ar, return_index, return_inverse, return_counts)
224 if not (-ar.ndim <= axis < ar.ndim):
225 raise ValueError('Invalid axis kwarg specified for unique')
c:\users\hambarsous\appdata\local\programs\python\python36\lib\site-packages\numpy\lib\arraysetops.py in _unique1d(ar, return_index, return_inverse, return_counts)
278
279 if optional_indices:
--> 280 perm = ar.argsort(kind='mergesort' if return_index else 'quicksort')
281 aux = ar[perm]
282 else:
TypeError: '<' not supported between instances of 'str' and 'int'
I've tried things along the lines of the following to see if I could get around the type error but no luck.
import pandas as pd
pd.to_numeric(data['value'], errors='coerce')
Below is the code that produces my error, my only assumption at this point is when I read the data in and melt it maybe that's doing something to the data type that I am not fully understanding. When I check the data types for value and sample_id I get float64 and object.
import pandas as pd
data = pd.read_excel('CHOK1CGA2016WB1.xlsx', sheet_name='Cleaned Data')
data = pd.melt(data, id_vars=['sample_code', 'sample_id', 'day', 'test'], value_vars=['rep 1', 'rep 2', 'rep 3', 'rep 4', 'rep 5', 'rep 6'])
data.rename(columns={'variable': 'rep'}, inplace=True)
from statsmodels.stats.multicomp import pairwise_tukeyhsd
pairwise_tukeyhsd(endog=data['value'], groups=data['sample_id'])
The comment to my question put me on the path of further inspecting the elements of the data['sample_id'] column. It seems this is where the problem lied, since it was a mix of number or letter/numbers I believe the number only entries were causing the error. I simply added some letters into the elements that had numbers and the code ran.