Refactor and @njit on methods

[fjloyola]

Refactor and @njit on methods:

• adding @njit in the methods get_new_idx_data_points and draw_leaf_value;
• avoid creating variables when it is posible;
• change the get_split_variables() to be a generator;
• instead of creating Particle and then put the value of the kfactor we refactor constructor to have the default value in 0.75 as a parameter;
• grow_tree now returns the new leaf idx;
• in grow_tree, change the if else sentence for early return, also, instead of creating new node split node and delete the leaf value there is a new method grow_leaf_node;
• in the draw_leaf_value we simplify the expresions;
• in _travearse_tree instead of having two returns we create a variable next_node;

Tested Mac with chip M1, 8 cores and 16 gb of ram having the next results:

Configuration			version performances		improvements
model name	tree º	particle º	current	new
biking	50	20	74 sec	68 sec	8,11%
		40	99 sec	93 sec	6,06%
		60	128 sec	121 sec	5,47%
	100	20	103 sec	96 sec	6,80%
		40	154 sec	140 sec	9,09%
		60	209 sec	200 sec	4,31%
	200	20	164 sec	155 sec	5,49%
		40	267 sec	255 sec	4,49%
		60	378 sec	359 sec	5,03%
coal	50	20	57 sec	57 sec	0,00%
		40	83 sec	81 sec	2,41%
		60	104 sec	94 sec	9,61%
	100	20	85 sec	77 sec	9,41%
		40	135 sec	108 sec	20,00%
		60	181 sec	145 sec	19,89%
	200	20	138 sec	128 sec	7,24%
		40	229 sec	203 sec	11,35%
		60	319 sec	278 sec	12,85%
friedman	50	20	65 sec	57 sec	12,31%
		40	91 sec	75 sec	17,58%
		60	115 sec	96 sec	16,52%
	100	20	93 sec	79 sec	15,05%
		40	144 sec	118 sec	18,06%
		60	200 sec	162 sec	19,00%
	200	20	150 sec	125 sec	16,67%
		40	246 sec	202 sec	17,89%
		60	354 sec	279 sec	21,19%
space_influenza	50	20	66 sec	59 sec	10,61%
		40	95 sec	77 sec	18,95%
		60	125 sec	101 sec	19,20%
	100	20	101 sec	81 sec	17,65%
		40	153 sec	126 sec	17,65%
		60	207 sec	171 sec	17,39%
	200	20	158 sec	134 sec	15,19%
		40	271 sec	220 sec	18,82%
		60	381 sec	313 sec	17,85%