gui tests

Author: bartzbeielstein

gui/spotGUI_1.py

@@ -1,21 +1,17 @@
 import tkinter as tk
 import json
 from spotPython.hyperparameters.values import get_default_values, get_bound_values
+from spotPython.hyperdict.light_hyper_dict import LightHyperDict
 
-def get_default_values(elements):
-    pass
 
-def get_bound_values(elements):
-    pass
-
-def create_gui(dict_file):
-    # Load the dictionary from the file
-    with open(dict_file, 'r') as f:
-        elements = json.load(f)
+def create_gui(model):
+    lhd = LightHyperDict()
+    # generate a dictionary fun_control with the key "core_model_hyper_dict" and the value lhd.hyper_dict['NetLightRegression']
+    fun_control = {"core_model_hyper_dict": lhd.hyper_dict['NetLightRegression']}
 
     # Apply the functions to the dictionary
-    default_values = get_default_values(elements)
-    bound_values = get_bound_values(elements)
+    default_values = get_default_values(fun_control)
+    bound_values = get_bound_values(fun_control)
 
     # Create a tkinter window
     root = tk.Tk()
@@ -40,5 +36,5 @@ def create_gui(dict_file):
     root.mainloop()
 
 # Call the function to create the GUI
-create_gui("elements.json")
+create_gui(model = 'NetLightRegression')
 

gui/spotPythonGUI.py

@@ -0,0 +1,224 @@
+import tkinter as tk
+from tkinter import ttk
+
+from spotRiver.tuner.run import run_spot_river_experiment, compare_tuned_default, contour_plot, parallel_plot, importance_plot, progress_plot
+
+result = None
+fun_control = None
+
+
+def run_experiment():
+    global result, fun_control
+    MAX_TIME = float(max_time_entry.get())
+    INIT_SIZE = int(init_size_entry.get())
+    PREFIX = prefix_entry.get()
+    horizon = int(horizon_entry.get())
+    n_total = n_total_entry.get()
+    if n_total == "None" or n_total == "All":
+        n_total = None
+    else:
+        n_total = int(n_total)
+    perc_train = float(perc_train_entry.get())
+    oml_grace_period = oml_grace_period_entry.get()
+    if oml_grace_period == "None" or oml_grace_period == "n_train":
+        oml_grace_period = None
+    else:
+        oml_grace_period = int(oml_grace_period)
+    data_set = data_set_combo.get()
+    prep_model = prep_model_combo.get()
+    core_model = core_model_combo.get()
+
+    result, fun_control = run_spot_river_experiment(
+        MAX_TIME=MAX_TIME,
+        INIT_SIZE=INIT_SIZE,
+        PREFIX=PREFIX,
+        horizon=horizon,
+        n_total=n_total,
+        perc_train=perc_train,
+        oml_grace_period=oml_grace_period,
+        data_set=data_set,
+        prepmodel=prep_model,
+        coremodel=core_model,
+        log_level=20,
+    )
+
+
+def call_compare_tuned_default():
+    if result is not None and fun_control is not None:
+        compare_tuned_default(result, fun_control)
+
+
+def call_parallel_plot():
+    if result is not None:
+        parallel_plot(result)
+
+
+def call_contour_plot():
+    if result is not None:
+        contour_plot(result)
+
+
+def call_importance_plot():
+    if result is not None:
+        importance_plot(result)
+
+
+def call_progress_plot():
+    if result is not None:
+        progress_plot(result)
+
+
+# Create the main application window
+app = tk.Tk()
+app.title("Spot River Hyperparameter Tuning GUI")
+
+# Create a notebook (tabbed interface)
+notebook = ttk.Notebook(app)
+# notebook.pack(fill='both', expand=True)
+
+# Create and pack entry fields for the "Run" tab
+run_tab = ttk.Frame(notebook)
+notebook.add(run_tab, text="Binary classification")
+
+# colummns 0+1: Data
+
+data_label = tk.Label(run_tab, text="Data options:")
+data_label.grid(row=0, column=0, sticky="W")
+
+data_set_label = tk.Label(run_tab, text="Select data_set:")
+data_set_label.grid(row=1, column=0, sticky="W")
+data_set_values = ["Bananas", "CreditCard", "Elec2", "Higgs", "HTTP", "MaliciousURL", "Phishing", "SMSSpam", "SMTP", "TREC07", "USER"]
+data_set_combo = ttk.Combobox(run_tab, values=data_set_values)
+data_set_combo.set("Phishing")  # Default selection
+data_set_combo.grid(row=1, column=1)
+
+
+n_total_label = tk.Label(run_tab, text="n_total:")
+n_total_label.grid(row=2, column=0, sticky="W")
+n_total_entry = tk.Entry(run_tab)
+n_total_entry.insert(0, "All")
+n_total_entry.grid(row=2, column=1, sticky="W")
+
+perc_train_label = tk.Label(run_tab, text="perc_train:")
+perc_train_label.grid(row=3, column=0, sticky="W")
+perc_train_entry = tk.Entry(run_tab)
+perc_train_entry.insert(0, "0.60")
+perc_train_entry.grid(row=3, column=1, sticky="W")
+
+
+# colummns 2+3: Model
+model_label = tk.Label(run_tab, text="Model options:")
+model_label.grid(row=0, column=2, sticky="W")
+
+prep_model_label = tk.Label(run_tab, text="Select preprocessing model")
+prep_model_label.grid(row=1, column=2, sticky="W")
+prep_model_values = ["MinMaxScaler", "StandardScaler", "None"]
+prep_model_combo = ttk.Combobox(run_tab, values=prep_model_values)
+prep_model_combo.set("StandardScaler")  # Default selection
+prep_model_combo.grid(row=1, column=3)
+
+
+core_model_label = tk.Label(run_tab, text="Select core model")
+core_model_label.grid(row=2, column=2, sticky="W")
+core_model_values = ["AMFClassifier", "HoeffdingAdaptiveTreeClassifier", "LogisticRegression"]
+core_model_combo = ttk.Combobox(run_tab, values=core_model_values)
+core_model_combo.set("LogisticRegression")  # Default selection
+core_model_combo.grid(row=2, column=3)
+
+
+# columns 4+5: Experiment
+experiment_label = tk.Label(run_tab, text="Experiment options:")
+experiment_label.grid(row=0, column=4, sticky="W")
+
+max_time_label = tk.Label(run_tab, text="MAX_TIME:")
+max_time_label.grid(row=1, column=4, sticky="W")
+max_time_entry = tk.Entry(run_tab)
+max_time_entry.insert(0, "1")
+max_time_entry.grid(row=1, column=5)
+
+init_size_label = tk.Label(run_tab, text="INIT_SIZE:")
+init_size_label.grid(row=2, column=4, sticky="W")
+init_size_entry = tk.Entry(run_tab)
+init_size_entry.insert(0, "3")
+init_size_entry.grid(row=2, column=5)
+
+prefix_label = tk.Label(run_tab, text="PREFIX:")
+prefix_label.grid(row=3, column=4, sticky="W")
+prefix_entry = tk.Entry(run_tab)
+prefix_entry.insert(0, "00")
+prefix_entry.grid(row=3, column=5)
+
+horizon_label = tk.Label(run_tab, text="horizon:")
+horizon_label.grid(row=4, column=4, sticky="W")
+horizon_entry = tk.Entry(run_tab)
+horizon_entry.insert(0, "1")
+horizon_entry.grid(row=4, column=5)
+
+oml_grace_period_label = tk.Label(run_tab, text="oml_grace_period:")
+oml_grace_period_label.grid(row=5, column=4, sticky="W")
+oml_grace_period_entry = tk.Entry(run_tab)
+oml_grace_period_entry.insert(0, "n_train")
+oml_grace_period_entry.grid(row=5, column=5)
+
+# column 6: Run button
+run_button = ttk.Button(run_tab, text="Run Experiment", command=run_experiment)
+run_button.grid(row=7, column=6, columnspan=2, sticky="E")
+
+# Create and pack the "Regression" tab with a button to run the analysis
+regression_tab = ttk.Frame(notebook)
+notebook.add(regression_tab, text="Regression")
+
+# colummns 0+1: Data
+
+regression_data_label = tk.Label(regression_tab, text="Data options:")
+regression_data_label.grid(row=0, column=0, sticky="W")
+
+# colummns 2+3: Model
+regression_model_label = tk.Label(regression_tab, text="Model options:")
+regression_model_label.grid(row=0, column=2, sticky="W")
+
+# columns 4+5: Experiment
+regression_experiment_label = tk.Label(regression_tab, text="Experiment options:")
+regression_experiment_label.grid(row=0, column=4, sticky="W")
+
+
+# Create and pack the "Analysis" tab with a button to run the analysis
+analysis_tab = ttk.Frame(notebook)
+notebook.add(analysis_tab, text="Analysis")
+
+notebook.pack()
+
+
+# Add the Logo image in both tabs
+logo_image = tk.PhotoImage(file="images/spotlogo.png")
+logo_label = tk.Label(run_tab, image=logo_image)
+logo_label.grid(row=0, column=6, rowspan=1, columnspan=1)
+
+analysis_label = tk.Label(analysis_tab, text="Analysis options:")
+analysis_label.grid(row=0, column=1, sticky="W")
+
+progress_plot_button = ttk.Button(analysis_tab, text="Progress plot", command=call_progress_plot)
+progress_plot_button.grid(row=1, column=1, columnspan=2, sticky="W")
+
+compare_tuned_default_button = ttk.Button(analysis_tab, text="Compare tuned vs. default", command=call_compare_tuned_default)
+compare_tuned_default_button.grid(row=2, column=1, columnspan=2, sticky="W")
+
+importance_plot_button = ttk.Button(analysis_tab, text="Importance plot", command=call_importance_plot)
+importance_plot_button.grid(row=3, column=1, columnspan=2, sticky="W")
+
+contour_plot_button = ttk.Button(analysis_tab, text="Contour plot", command=call_contour_plot)
+contour_plot_button.grid(row=4, column=1, columnspan=2, sticky="W")
+
+parallel_plot_button = ttk.Button(analysis_tab, text="Parallel plot (Browser)", command=call_parallel_plot)
+parallel_plot_button.grid(row=5, column=1, columnspan=2, sticky="W")
+
+
+analysis_logo_label = tk.Label(analysis_tab, image=logo_image)
+analysis_logo_label.grid(row=0, column=6, rowspan=1, columnspan=1)
+
+regression_logo_label = tk.Label(regression_tab, image=logo_image)
+regression_logo_label.grid(row=0, column=6, rowspan=1, columnspan=1)
+
+# Run the mainloop
+
+app.mainloop()