feature_matrix_train: Training feature matrix
In ss15859/tennis.wta: Predicts Tennis Matches

Description Usage Arguments Value Examples

Makes the feature matrix to train on based on which features are chosen. Uses functions: player_data_frame, head_to_head, feature_surface.

1	feature_matrix_train(my_data,match_date = "2019-01-01",features = c('surface_form', 'age', 'match_form', 'head_to_head'), carpet = 'FALSE')

`my_data`	input data.
`match_date`	set to the starting date of test data - by default this is 2019-01-01.
`features`	a list of features wanted to be included. By default all features are used.
`carpet`	as matches are no longer played on carpet, carpet performance is by default not included. However, carpet can be included by including, carpet == 'TRUE'

Creates a matrix where each row is a feature and each column is a match. Additionally, a row of ones is added for the intercept and a row which signifies the winner of that match - set to +1 if player 1 one and -1 if player 2 won.

1 2	feature_matrix_train(data) feature_matrix_train(my_data,match_date = "2018-05-11",features = c('surface_form', 'match_form', 'head_to_head'), carpet = 'TRUE')