Initialized projec with code from @nrutkowski1

Author: MatthewSpofford

CanvasPanel.py

@@ -0,0 +1,95 @@
+
+import numpy as np
+from scipy.optimize import curve_fit
+
+# Class CurveFit:
+# contains all functions used to generate a regression line given a set of data points
+# functions are called in R2byScalePlot and RegressionPlot class functions.
+class CurveFit:
+
+    def __init__(self):
+
+        # Regression line is determined recursively using the curve_fit function from scipy library.
+        # this value can be changed in the Curve Fit dialog by the user. I set the default to be 1000 based on trial and error.
+        # this can actually have a significant impact with R^2 values by changing them from undefined / zero all the way to 0.3+
+        # depending on the scale
+        self.maxfev = 1000
+
+    # below are the functions for each curve type easy to add new functions or in the future have user defined functions
+    # -----------------------------Proportional Fit Functions---------------------------------------------------------------
+    def prop_fit(self, x, a): return a*x
+    def prop_data(self, x, y): return curve_fit(self.prop_fit, x, y, maxfev=self.get_maxfev())
+    # ------------------------------Linear Fit Functions--------------------------------------------------------------------
+    def linear_fit(self, x, a, b): return a*x + b
+    def linear_data(self, x, y): return curve_fit(self.linear_fit, x, y, maxfev=self.get_maxfev())
+    # ------------------------------Quadratic Fit Functions-----------------------------------------------------------------
+    def quad_fit(self, x, a, b, c): return a*x**2 + b*x + c
+    def quad_data(self, x, y): return curve_fit(self.quad_fit, x, y, maxfev=self.get_maxfev())
+    # ------------------------------Cubic Fit Functions---------------------------------------------------------------------
+    def cubic_fit(self, x, a, b, c, d): return a*x**3 + b*x**2 + c*x + d
+    def cubic_data(self, x, y): return curve_fit(self.cubic_fit, x, y, maxfev=self.get_maxfev())
+    # ---------------------------------Quartic Fit Functions----------------------------------------------------------------
+    def quartic_fit(self, x, a, b, c, d, e): return a*x**4 + b*x**3 + c*x**2 + d*x + e
+    def quartic_data(self, x, y): return curve_fit(self.quartic_fit, x, y, maxfev=self.get_maxfev())
+    # ------------------------------- Quintic Fit Functions-----------------------------------------------------------------
+    def quintic_fit(self, x, a, b, c, d, e, f): return a*x**5 + b*x**4 + c*x**3 + d*x**2 + e*x + f
+    def quintic_data(self, x, y): return curve_fit(self.quintic_fit, x, y, maxfev=self.get_maxfev())
+    # -------------------------------------Power Fit Functions -----------------------------------------------------------
+    def power_fit(self, x, a, b): return a*x**b
+    def power_data(self, x, y): return curve_fit(self.power_fit, x, y, maxfev=self.get_maxfev())
+    # -------------------------------------Inverse Fit Functions-------------------------------------------------------
+    def inverse_fit(self, x, a): return a/x
+    def inverse_data(self, x, y): return curve_fit(self.inverse_fit, x, y, maxfev=self.get_maxfev())
+    # ---------------------------------------Inverse Squared Fit Functions -------------------------------------------------
+    def insq_fit(self, x, a): return a/(x**2)
+    def insq_data(self, x, y): return curve_fit(self.insq_fit, x, y, maxfev=self.get_maxfev())
+    # ---------------------------------------------Natural Exponent Fit Functions-------------------------------------------
+    def nexp_fit(self, x, a, b, c): return a*np.exp(-1*b*x) + c
+    def nexp_data(self, x, y): return curve_fit(self.nexp_fit, x, y, maxfev=self.get_maxfev())
+    # ----------------------------------------Natural Log Fit--------------------------------------------------------------
+    def ln_fit(self, x, a, b): return a*np.log(b*x)  # second term is base in log()
+    def ln_data(self, x, y): return curve_fit(self.ln_fit, x, y, maxfev=self.get_maxfev())
+    # --------------------------------------Base-10 Exponent----------------------------------------------------------------
+    # doesnt even work smh
+    def b10exp_fit(self, x, a, b, c): return a*(10**(b*x)) + c
+    def b10exp_data(self, x, y): return curve_fit(self.b10exp_fit, x, y, maxfev=self.get_maxfev())
+    # ---------------------------------------Base-10 Logarithm--------------------------------------------------------------
+    def b10log_fit(self, x, a, b): return a*np.log10(b*x)
+    def b10log_data(self, x, y): return curve_fit(self.b10log_fit, x, y, maxfev=self.get_maxfev())
+    # ---------------------------------------Inverse Exponent Fit-----------------------------------------------------------
+    def invexp_fit(self, x, a, b, c): return a*(1-np.exp(-1*b*x)) + c
+    def invexp_data(self, x, y): return curve_fit(self.invexp_fit, x, y, maxfev=self.get_maxfev())
+    # ---------------------------------------Sine Fit------------------------------------------------------------------
+    def sine_fit(self, x, a, b, c, d): return a*np.sin(b*x + c) + d
+    def sine_data(self, x, y): return curve_fit(self.sine_fit, x, y, maxfev=self.get_maxfev())
+    # -------------------------------------Cosine Fit----------------------------------------------------
+    def cosine_fit(self, x, a, b, c, d): return a*np.cos(b*x + c) + d
+    def cosine_data(self, x, y): return curve_fit(self.cosine_fit, x, y, maxfev=self.get_maxfev())
+    # -------------------------------------Cosine Squared Fit--------------------------------------------------------
+    # not sure if theres a point to this it doesnt even work smh
+    def cossqrd_fit(self, x, a, b, c, d): return a*np.square(np.cos(b*x + c)) + d
+    def cossqrd_data(self, x, y): return curve_fit(self.cossqrd_fit, x, y, maxfev=self.get_maxfev())
+    # ---------------------------------- Gaussian Fit ----------------------------------------------------------------
+    def gauss_fit(self, x, a, b, c, d): return a*np.exp(-1*((x-b)**2)/(c**2)) + d
+    def gauss_data(self, x, y): return curve_fit(self.gauss_fit, x, y, maxfev=self.get_maxfev())
+    # ---------------------------------------R^2 CALCULATION----------------------------------------------------------------
+    # function to calculate the R^2 value of a regression line
+    def r_squared(self, y, func):
+
+        # forumla for R^2 calculation
+        residuals = y - np.array(func)
+        # residual sum of squares
+        ss_res = np.sum(residuals**2)
+        ss_total = np.sum((y - np.mean(y))**2)
+
+        # depending on the data, the regression function can yield results > 1 or ss_total = 0 so R^2 would be undefined
+        # this returns nan so in each of these cases the R^2 value is set to be 0
+        if ss_total == 0:
+            return 0
+        elif (ss_res / ss_total) > 1:
+            return 0
+        else:
+            return np.sqrt(1 - (ss_res / ss_total))
+
+    def get_maxfev(self): return self.maxfev
+    def set_maxfev(self, maxfev): self.maxfev = maxfev