Add nodes, leaves, depth to mysql

Add nodes, leaves, depth, samples, features and classes to analysis

analysis_mysql.py

@@ -1,5 +1,6 @@
 import argparse
 from typing import Tuple
+import numpy as np
 from experimentation.Sets import Datasets
 from experimentation.Utils import TextColor
 from experimentation.Database import MySQL
@@ -14,8 +15,10 @@ models_tree = [
     "baseRaF",
 ]
 models_ensemble = ["odte", "adaBoost", "bagging", "TBRaF", "TBRoF", "TBRRoF"]
+description = ["samp", "var", "cls"]
+complexity = ["nodes", "leaves", "depth"]
 title = "Best model results"
-lengths = (30, 12, 12, 12, 12, 12, 12)
+lengths = (30, 4, 3, 3, 3, 3, 3, 12, 12, 12, 12, 12, 12)
 
 
 def parse_arguments() -> Tuple[str, str, str, bool, bool]:
@@ -79,9 +82,11 @@ def report_header(title, experiment, model_type):
 
 def report_line(line):
     output = f"{line['dataset']:{lengths[0] + 5}s} "
+    for key, item in enumerate(description + complexity):
+        output += f"{line[item]:{lengths[key + 1]}d} "
     data = models.copy()
     for key, model in enumerate(data):
-        output += f"{line[model]:{lengths[key + 1]}s} "
+        output += f"{line[model]:{lengths[key + 7]}s} "
     return output
 
 
@@ -101,7 +106,15 @@ def report_footer(agg):
 dbh = MySQL()
 database = dbh.get_connection()
 dt = Datasets(False, False, "tanveer")
-fields = ("Dataset",)
+fields = (
+    "Dataset",
+    "Samp",
+    "Var",
+    "Cls",
+    "Nod",
+    "Lea",
+    "Dep",
+)
 models = models_tree if model_type == "tree" else models_ensemble
 for item in models:
     fields += (f"{item}",)
@@ -121,13 +134,23 @@ for dataset in dt:
     find_one = False
     # Look for max accuracy for any given dataset
     line = {"dataset": color + dataset[0]}
+    X, y = dt.load(dataset[0])  # type: ignore
+    line["samp"], line["var"] = X.shape
+    line["cls"] = len(np.unique(y))
     record = dbh.find_best(dataset[0], models, experiment)
     max_accuracy = 0.0 if record is None else record[5]
+    line["nodes"] = 0
+    line["leaves"] = 0
+    line["depth"] = 0
     for model in models:
         record = dbh.find_best(dataset[0], model, experiment)
         if record is None:
             line[model] = color + "-" * 12
         else:
+            if model == "stree":
+                line["nodes"] = record[12]
+                line["leaves"] = record[13]
+                line["depth"] = record[14]
             reference = record[13]
             accuracy = record[5]
             acc_std = record[11]