[FIXED] How to remove rows from a specific column for preparing training and test set for SVR

Issue

A dataset has more than 2500 rows and 22 columns including the age column. I have completed all of the processes for SVR. It going on. But I am still having to face an error. That is raise ValueError("bad input shape {0}".format(shape)), ValueError: bad input shape (977, 57). My input is SupportVectorRefModel.fit(X_train, y_train). How can I resolve this problem?

from sklearn.model_selection 
import train_test_split 
from sklearn.svm import SVR 

X_train, y_train = dataset.loc[:1000], dataset.loc[:1000] 
X_test, y_test = dataset.loc[1001], dataset.loc[1001] 
train_X, train_y = X_train.drop(columns=['age']), y_train.pop('age')
test_X, test_y = X_test.drop(columns=['age']), y_test.pop('age')

raise ValueError("bad input shape {0}".format(shape))
ValueError: bad input shape (977, 57)

Solution

if you want to drop the first x rows of a given column, useDataFrame.drop

df = df.drop(range(end))

Where end is the number of rows you want to remove from the beginning of the dataset.

Note: you cannot drop elements from a single column only. You must either drop all rows or replace the removed values with NaN or some other replacement value

Update

After further clarification by OP, the end result is to remove the agecolumn from the X_* dataframes, and store the age column in its own dataframe *_y

from sklearn.model_selection 
import train_test_split 
from sklearn.svm import SVR 

X_train, y_train = dataset.loc[:1000], dataset.loc[:1000] 
X_test, y_test = dataset.loc[1001], dataset.loc[1001] 
train_X, train_y = X_train.drop(columns=['age']), y_train.pop('age')
test_X, test_y = X_test.drop(columns=['age']), y_test.pop('age')

Should give you the desired end result.

Answered By - Bobs Burgers