PEP8

safep/TI.py

@@ -4,15 +4,15 @@
 import numpy as np
 
 
-def processTI(dataTI, restraint, Lsched):
+def process_TI(dataTI, restraint, Lsched):
     '''
     Arguments: the TI data, restraint, and lambda schedule
     Function: Calculate the free energy for each lambda value, aggregate the result, and estimate the error
     Returns: The free energies and associated errors as functions of lambda. Both per window and cumulative.
     '''
     dUs = {}
     for key, group in dataTI.groupby('L'):
-        dUs[key] = [HW_dUdL(restraint, coord, key) for coord in group.DBC]
+        dUs[key] = [harmonicWall_dUdL(restraint, coord, key) for coord in group.DBC]
 
     Lsched = np.sort(list(dUs.keys()))
     dL = Lsched[1] - Lsched[0]
@@ -27,7 +27,7 @@ def processTI(dataTI, restraint, Lsched):
     return TIperWindow, TIcumulative
 
 
-def plotTI(cumulative, perWindow, width=8, height=4, PDFtype='KDE', hystLim=(-1,1), color='#0072B2'):
+def plot_TI(cumulative, perWindow, width=8, height=4, PDFtype='KDE', hystLim=(-1,1), color='#0072B2'):
     fig, (cumAx,eachAx) = plt.subplots(2,1, sharex='col')
 
     # Cumulative change in kcal/mol
@@ -47,11 +47,11 @@ def plotTI(cumulative, perWindow, width=8, height=4, PDFtype='KDE', hystLim=(-1,
 
     return fig, [cumAx,eachAx] 
 
-def makeHarmonicWall(FC=10, targetFC=0, targetFE=1, upperWalls=1, schedule=None, numSteps=1000, targetEQ=500, name='HW', lowerWalls=None):
+def make_harmonicWall(FC=10, targetFC=0, targetFE=1, upperWalls=1, schedule=None, numSteps=1000, targetEQ=500, name='HW', lowerWalls=None):
     HW = {'name':name, 'targetFC':targetFC, 'targetFE':targetFE, 'FC':FC, 'upperWalls':upperWalls, 'schedule':schedule, 'numSteps':numSteps, 'targetEQ':targetEQ, 'lowerWalls':lowerWalls}
     return HW
 
-def HW_U(HW, coord, L):
+def harmonicWall_U(HW, coord, L):
     d=0
     if HW['upperWalls'] and coord>HW['upperWalls']:
         d = coord-HW['upperWalls']
@@ -67,7 +67,7 @@ def HW_U(HW, coord, L):
         U=0
     return U
 
-def HW_dUdL(HW, coord, L):
+def harmonicWall_dUdL(HW, coord, L):
     d=0
     if HW['upperWalls'] and coord>HW['upperWalls']:
         d = coord-HW['upperWalls']

safep/estimators.py

@@ -27,7 +27,7 @@
 from .helpers import *
 
 
-def get_EXP(u_nk):
+def get_exponential(u_nk):
     '''
     Get exponential estimation of the change in free energy.
     Arguments: u_nk in alchemlyb format
@@ -96,7 +96,7 @@ def do_estimation(u_nk, method='both'):
         perWindow.index = l_mids
 
     if method=='both' or method=='EXP':
-        expl, expmid, dG_fs, dG_bs = get_EXP(u_nk)
+        expl, expmid, dG_fs, dG_bs = get_exponential(u_nk)
 
         cumulative[('EXP', 'ff')] = np.insert(np.cumsum(dG_fs),0,0)
         cumulative[('EXP', 'fb')] = np.insert(-np.cumsum(dG_bs),0,0)
@@ -117,7 +117,7 @@ def do_estimation(u_nk, method='both'):
 
 
 #Light-weight exponential estimator. Requires alternative parser.
-def get_dG_fromData(data, temperature):
+def get_dG_from_data(data, temperature):
     from scipy.constants import R, calorie
     beta = 1/(R/(1000*calorie) * temperature) #So that the final result is in kcal/mol
 

safep/fileIO.py

@@ -27,7 +27,7 @@
 from .helpers import *
 
 
-def guessLambda(fname):
+def guess_lambda(fname):
     '''
     Guess lambda based on file name (last number in the filename divided by 100). Not very good.
     Arguments: file name
@@ -36,14 +36,14 @@ def guessLambda(fname):
     L = int(re.findall(r'\d+', fname)[-1])/100
     return L
 
-def saveUNK(u_nk, filepath):
+def save_UNK(u_nk, filepath):
     '''
     Write u_nk to a file
     Arguments: u_nk in the format of alchemlyb, filepath
     '''
     u_nk.to_csv(filepath)
 
-def readUNK(filepath):
+def read_UNK(filepath):
     '''
     Read a u_nk that was written by saveUNK.
     Arguments: filepath
@@ -56,7 +56,7 @@ def readUNK(filepath):
     return u_nk.copy()
 
 
-def readFEPOUT(fileName, step=1):
+def read_FEPOUT(fileName, step=1):
     '''
     Reads all data from a NAMD fepout file unlike alchemlyb's parser.
     readFEPOUT reads each file in a single pass: keeping track of lambda values and appending each line to an array. 
@@ -93,7 +93,7 @@ def readFEPOUT(fileName, step=1):
             elif frame % step <= 1:
                 if np.isnan(L):
                     print("WARNING: lambda is not defined!")
-                    L = guessLambda(fileName)
+                    L = guess_lambda(fileName)
                     print("Guessing lambda to be {L} based on file name.")
 
 
@@ -126,15 +126,15 @@ def readFEPOUT(fileName, step=1):
     df = df.sort_index()
     return df
 
-def readFiles(files, step=1):
+def read_Files(files, step=1):
     '''
     Batch readFEPOUT
     Arguments: file (paths), step (stride)
     Returns: a list of dataframes, one dataframe per file
     '''
     fileList = []
     for file in files:
-        df = readFEPOUT(file, step)
+        df = read_FEPOUT(file, step)
         fileList.append(df)
     data = pd.concat(fileList)
 

safep/helpers.py

@@ -25,7 +25,7 @@
 from glob import glob #file regexes
 
 # Calculate the coefficient of determination:
-def GetRsq(X, Y, Yexpected):
+def get_Rsq(X, Y, Yexpected):
     '''
     Calculate the coefficient of determination for arbitrary fits.
     Arguments: X (inputs), Y (experimental data), Yexpected (fitted or predicted data)
@@ -37,14 +37,14 @@ def GetRsq(X, Y, Yexpected):
     return R
 
 #Wrappers
-def cumFn(x, m, s):
+def cum_Fn(x, m, s):
     '''
     Wrapper for the normal cumulative density function
     '''
     r = norm.cdf(x, m, s)
     return r
 
-def pdfFn(x,m,s):
+def pdf_Fn(x,m,s):
     '''
     Wrapper for the normal probability density function
     '''

safep/plotting.py

@@ -10,6 +10,7 @@
 from scipy.special import erfc
 from scipy.optimize import curve_fit as scipyFit
 from scipy.stats import skew
+import scipy as sp
 
 import pandas as pd
 
@@ -27,7 +28,7 @@
 from .helpers import *
 
 
-def convergencePlot(theax, fs, ferr, bs, berr, fwdColor='#0072B2', bwdColor='#D55E00', lgndF=None, lgndB=None):
+def convergence_plot(theax, fs, ferr, bs, berr, fwdColor='#0072B2', bwdColor='#D55E00', lgndF=None, lgndB=None):
     '''
     A revised convergence plot. Plays nicely with other plotting functions and is more customizable.
     Arguments: theax[is] on which to plot, fs (forward estimates), ferr (forward errors), bs (backward-sampled estimates), berr (backward-sampled errors), fwdColor, bwdColor, lgndF legend forward color, lgndB legend backward color
@@ -49,7 +50,7 @@ def convergencePlot(theax, fs, ferr, bs, berr, fwdColor='#0072B2', bwdColor='#D5
 
     return theax
 
-def doConvPlot(ax, X, fs, ferr, fwdColor, label=None):
+def do_conv_plot(ax, X, fs, ferr, fwdColor, label=None):
     '''
     A minimalist convergence plot
     Arguments: ax[is], X (the X values), fs (forward estimates), ferr (forward errors), fwdColor (the forward color), label text
@@ -58,29 +59,8 @@ def doConvPlot(ax, X, fs, ferr, fwdColor, label=None):
     ax.errorbar(X, fs, yerr=ferr, marker=None, linewidth=1, color=fwdColor, markerfacecolor='white', markeredgewidth=1, markeredgecolor=fwdColor, ms=5, label=label)
     return ax
 
-#Cannonical convergence plot
-def convergence_plot(u_nk, tau=1, units='kT', RT=0.59):
-    '''
-    Older convergence plot. Does the convergence calculation and plotting. Deprecated.
-    Arguments: u_nk, tau (an error tuning factor), units (kT or kcal/mol), RT
-    Returns: a pyplot
-    '''
-    forward, forward_error, backward, backward_error = doConvergence(u_nk, num_points=10)
-
-    if units=='kcal/mol':
-        forward = forward*RT
-        forward_error = forward_error*RT
-        backward = backward*RT
-        backward_error = backward_error*RT
-
-    ax = plot_convergence(forward, forward_error, backward, backward_error)
-
-    if units=='kcal/mol':
-        ax.set(ylabel=r'$\rm\Delta G$'+'\n(kcal/mol)')
-
-    return plt.gca()
 
-def convergencePlot(theax, fs, ferr, bs, berr, fwdColor='#0072B2', bwdColor='#D55E00', lgndF=None, lgndB=None):
+def convergence_plot(theax, fs, ferr, bs, berr, fwdColor='#0072B2', bwdColor='#D55E00', lgndF=None, lgndB=None):
     '''
     Convergence plot. Does the convergence calculation and plotting.
     Arguments: u_nk, tau (an error tuning factor), units (kT or kcal/mol), RT
@@ -138,7 +118,7 @@ def fb_discrepancy_hist(dG_f, dG_b):
     return plt.gca()
 
 
-def plotGeneral(cumulative, cumulativeYlim, perWindow, perWindowYlim, RT, width=8, height=4, PDFtype='KDE'):
+def plot_general(cumulative, cumulativeYlim, perWindow, perWindowYlim, RT, width=8, height=4, PDFtype='KDE'):
     fig, ((cumAx, del1),( eachAx, del2), (ddGAx, del3), (hystAx, pdfAx)) = plt.subplots(4,2, sharex='col', sharey='row', gridspec_kw={'width_ratios': [2, 1]})
 
     fig.delaxes(del1)