y中人口最少的类只有1个成员，这太少了。 任何班级的最小团体人数不得少于2

Question

I am getting following error when I am trying to split data into train and test. I know that error is occurring because for stratify parameter I should pass categorical data only, not numeric but here OFFENSE_CODE is like a category except the categories in it are represented by number. So how can I do stratify sampling by OFFENSE_CODE .

x = df.loc[:,['YEAR','MONTH','DAY_OF_WEEK']]
X_train, x_test, Y_train, y_test = model_selection.train_test_split(x,df['OFFENSE_CODE'],stratify=df['OFFENSE_CODE'],random_state=2,test_size=0.3)

this is sample of dataset

INCIDENT_NUMBER  OFFENSE_CODE               OFFENSE_CODE_GROUP  \
  I192067438           613                          Larceny   
  I192067437          3831  Motor Vehicle Accident Response   
  I192067435          3115               Investigate Person   
  I192067434          3301                  Verbal Disputes   
  I192067433          3301                  Verbal Disputes   

                 OFFENSE_DESCRIPTION DISTRICT REPORTING_AREA SHOOTING  \
                LARCENY SHOPLIFTING       A1            112      NaN   
                    PROPERTY DAMAGE       A1                     NaN   
                 INVESTIGATE PERSON      C11            336      NaN   
                     VERBAL DISPUTE      E18            492      NaN   
                     VERBAL DISPUTE      D14            769      NaN   

  OCCURRED_ON_DATE  YEAR  MONTH DAY_OF_WEEK  HOUR    UCR_PART  \
2019-08-25 19:55:02  2019      8      Sunday    19    Part One   
2019-08-25 18:20:00  2019      8      Sunday    18  Part Three   
2019-08-25 20:45:00  2019      8      Sunday    20  Part Three   
2019-08-25 20:32:00  2019      8      Sunday    20  Part Three   
2019-08-25 20:30:00  2019      8      Sunday    20  Part Three   

      STREET        Lat       Long                     Location      CODES  
WASHINGTON ST  42.355123 -71.060880  (42.35512339, -71.06087980)   tyer613a  
        NaN  42.352389 -71.062603  (42.35238871, -71.06260312)  tyer3831a  
  NORTON ST  42.306265 -71.068646  (42.30626521, -71.06864556)  tyer3115a  
   DERRY RD  42.265933 -71.113774  (42.26593347, -71.11377415)  tyer3301a  
 PARSONS ST        NaN        NaN     (0.00000000, 0.00000000)  tyer3301a

i also tried

y = df.loc['OFFENSE_CODE'].apply(str)

X_train, x_test, Y_train, y_test = model_selection.train_test_split(x,y,stratify=y,random_state=2,test_size=0.3)

it is giving same error

ValueError:The least populated class in y has only 1 member, which is too few. The minimum number of groups for any class cannot be less than 2.

Answer 1

convert the column to string and then do the sampling

df['OFFENSE_CODE'].apply(str)

Dont foget to assign the result back

df['OFFENSE_CODE'] = df['OFFENSE_CODE'].apply(str)