fix a few images

TODO.md

@@ -5,17 +5,23 @@
 - [x] Use $E^f$ in Fig. 15 (and check discussion, especially **4**)
 - [x] Fig. 16: $E_{SHE}$ → $E^0_{abs}$
 - [x] Conclude & abstract & keywords
-- [x] xx in Fig. 12?
 - [x] Aknowledgements
 - [x] Clean up the bibliography, check for duplicates
 
-- [ ] pack the geometries
 - [ ] prepare a TOC
 - [ ] Research highlight
 - [ ] mark (and address) missing values in Tables
-- [ ] tries to move labels as much as possible in graphs
-- [ ] PACS code, MSC code?
-- [ ] Fig. or Figure?
-- [ ] Fig. S7: not $E^f$
-- [ ] Title of the SI
+- [x] Fig. S7: not $E^f$
+- [x] Title of the SI
 - [ ] Check NC → N-C+, etc.
+- [ ] xx in Fig. 12?
+- [ ] Use correct hammet parameters
+- [ ] Figs. S3 & S4
+- [ ] Entropy in cplx
+- [ ] Sign $U_q$
+- [ ] Extra explanation solv. (diff of $\varepsilon$ is not enough)
+
+- [ ] Fig. or Figure?
+- [ ] PACS code, MSC code?
+- [ ] tries to move labels as much as possible in graphs
+- [ ] pack the geometries

analyses/plot_cplx_Kx1.py

@@ -37,9 +37,9 @@ def plot_Kx1(ax, data: pandas.DataFrame, family: str, color: str):
 
     x = [int(x.replace('mol_', '')) for x in subdata['name']]
 
-    pK01 = numpy.log10(subdata['k01'])
-    pK11 = numpy.log10(subdata['k11'])
-    pK21 = numpy.log10(subdata['k21'])
+    pK01 = -numpy.log10(subdata['k01'])
+    pK11 = -numpy.log10(subdata['k11'])
+    pK21 = -numpy.log10(subdata['k21'])
 
     ax.plot(x, pK01, 'o', color=color, label=family.replace('Family.', ''))
     ax.plot(x, pK11, '^', color=color)
@@ -50,12 +50,12 @@ def plot_Kx1(ax, data: pandas.DataFrame, family: str, color: str):
 
 def plot_helpline(ax, data):
     x = [int(x.replace('mol_', '')) for x in data['name']]
-    ax.plot(x, numpy.log10(data['k01']), '--', color='black', linewidth=0.8)
 
-    pK01 = numpy.log10(data['k01'])
-    pK11 = numpy.log10(data['k11'])
-    pK21 = numpy.log10(data['k21'])
+    pK01 = -numpy.log10(data['k01'])
+    pK11 = -numpy.log10(data['k11'])
+    pK21 = -numpy.log10(data['k21'])
 
+    ax.plot(x, pK01, '--', color='black', linewidth=0.8)
     print('{} & {:.2f} $\\pm$ {:.2f} & {:.2f} $\\pm$ {:.2f} & {:.2f} $\\pm$ {:.2f} \\\\'.format('Total', numpy.mean(pK01), numpy.std(pK01), numpy.mean(pK11), numpy.std(pK11), numpy.mean(pK21), numpy.std(pK21)))
 
 def make_table(f, data: pandas.DataFrame, solvent: str):
@@ -101,7 +101,7 @@ def make_table(f, data: pandas.DataFrame, solvent: str):
 
 ax1.legend(ncols=5)
 ax1.set_xlim(0.5,61.5)
-ax1.text(38, 0.5, "Water", fontsize=18)
+ax1.text(38, 8, "Water", fontsize=18)
 ax1.xaxis.set_minor_locator(MultipleLocator(2))
 ax1.grid(which='both', axis='x')
 ax1.plot([0, 62], [0, 0], '-', color='grey')
@@ -116,12 +116,12 @@ def make_table(f, data: pandas.DataFrame, solvent: str):
 
 ax2.set_xlabel('Molecule id') 
 ax2.set_xlim(0.5, 61.5)
-ax2.text(38, 0.5, "Acetonitrile", fontsize=18)
+ax2.text(38, 8, "Acetonitrile", fontsize=18)
 ax2.xaxis.set_minor_locator(MultipleLocator(2))
 ax2.grid(which='both', axis='x')
 ax2.plot([0, 62], [0, 0], '-', color='grey')
 
-[ax.set_ylabel('log$_{10}$(K)') for ax in [ax1, ax2]]
+[ax.set_ylabel('pK$_{x1}$') for ax in [ax1, ax2]]
 
 plt.tight_layout()
 figure.savefig(args.output)

analyses/plot_cplx_Kx2.py

@@ -37,13 +37,13 @@ def plot_Kx2(ax, data: pandas.DataFrame, family: str, color: str):
 
     x = [int(x.replace('mol_', '')) for x in subdata['name']]
 
-    ax.plot(x, numpy.log10(subdata['k02']), 'o', color=color, label=family.replace('Family.', ''))
-    ax.plot(x, numpy.log10(subdata['k12']), '^', color=color)
-    ax.plot(x, numpy.log10(subdata['k22']), 's', color=color)
+    ax.plot(x, -numpy.log10(subdata['k02']), 'o', color=color, label=family.replace('Family.', ''))
+    ax.plot(x, -numpy.log10(subdata['k12']), '^', color=color)
+    ax.plot(x, -numpy.log10(subdata['k22']), 's', color=color)
 
 def plot_helpline(ax, data):
     x = [int(x.replace('mol_', '')) for x in data['name']]
-    ax.plot(x, numpy.log10(data['k02']), '--', color='black', linewidth=0.8)
+    ax.plot(x, -numpy.log10(data['k02']), '--', color='black', linewidth=0.8)
 
 def make_table(f, data: pandas.DataFrame, solvent: str):
     subdata = data[data['solvent'] == solvent]
@@ -88,12 +88,12 @@ def make_table(f, data: pandas.DataFrame, solvent: str):
 
 ax1.legend(ncols=5)
 ax1.set_xlim(0.5,61.5)
-ax1.text(38, -2, "Water", fontsize=18)
+ax1.text(38, 2, "Water", fontsize=18)
 ax1.xaxis.set_minor_locator(MultipleLocator(2))
 ax1.grid(which='both', axis='x')
 ax1.plot([0, 62], [0, 0], '-', color='grey')
 
-ax1.legend()
+ax1.legend(ncols=5)
 
 subdata_ac = prepare_data(data, 'acetonitrile')
 plot_helpline(ax2, subdata_ac)
@@ -105,12 +105,12 @@ def make_table(f, data: pandas.DataFrame, solvent: str):
 
 ax2.set_xlabel('Molecule id') 
 ax2.set_xlim(0.5,61.5)
-ax2.text(38, -2, "Acetonitrile", fontsize=18)
+ax2.text(38, 2, "Acetonitrile", fontsize=18)
 ax2.xaxis.set_minor_locator(MultipleLocator(2))
 ax2.grid(which='both', axis='x')
 ax2.plot([0, 62], [0, 0], '-', color='grey')
 
-[ax.set_ylabel('log$_{10}$(K)') for ax in [ax1, ax2]]
+[ax.set_ylabel('pK$_{x2}$') for ax in [ax1, ax2]]
 
 plt.tight_layout()
 figure.savefig(args.output)

analyses/plot_pot_er.py

@@ -10,10 +10,10 @@
 from nitroxides.commons import dG_DH, AU_TO_ANG, LabelPositioner, AU_TO_EV
 from nitroxides.tex import format_longtable
 
-LABELS = {'E_ox': [], 'E_red': []}
-POINTS_POSITION ={'E_ox': [], 'E_red': []}
-LABELS_KWARGS = {'E_ox': [], 'E_red': []}
-LABELS_PATH = {'E_ox': pathlib.Path('pot_er_ox.pos'), 'E_red': pathlib.Path('pot_er_red.pos')}
+LABELS = {'Ef_ox': [], 'Ef_red': []}
+POINTS_POSITION ={'Ef_ox': [], 'Ef_red': []}
+LABELS_KWARGS = {'Ef_ox': [], 'Ef_red': []}
+LABELS_PATH = {'Ef_ox': pathlib.Path('pot_er_ox.pos'), 'Ef_red': pathlib.Path('pot_er_red.pos')}
 
 AU_TO_DEBYE = 1 / 0.3934303 
 
@@ -23,42 +23,43 @@
     56, 58, # ?!?
 ]
 
-def plot_Er(ax, data: pandas.DataFrame, column: str, family: str, color: str):
-    subdata = data[(data['solvent'] == 'water') &  (data['family'] == family) & data['px'].notnull()]
+def prepare_data(data: pandas.DataFrame):
+    subdata = data[(data['solvent'] == 'water') & data['px'].notnull()]
+
+    dG_DH_ox = dG_DH(subdata['z'] + 1, subdata['z'], subdata['r_ox'] / AU_TO_ANG, subdata['r_rad'] / AU_TO_ANG, 80, 0) * AU_TO_EV
+    dG_DH_red = dG_DH(subdata['z'], subdata['z'] - 1, subdata['r_rad'] / AU_TO_ANG,  subdata['r_red'] / AU_TO_ANG, 80, 0) * AU_TO_EV
 
-    if column == 'E_ox':
-        dG_DH_ = dG_DH(subdata['z'] + 1, subdata['z'], subdata['r_ox'] / AU_TO_ANG, subdata['r_rad'] / AU_TO_ANG, 80, 0) * AU_TO_EV
-    else:
-        dG_DH_ = dG_DH(subdata['z'], subdata['z'] - 1, subdata['r_rad'] / AU_TO_ANG,  subdata['r_red'] / AU_TO_ANG, 80, 0) * AU_TO_EV
+    subdata.insert(1, 'Ef_ox', subdata['E_ox'] - dG_DH_ox)
+    subdata.insert(1, 'Ef_red', subdata['E_red'] - dG_DH_red)
 
-    Er = subdata['px'] / subdata['r'] ** 2 + subdata['Qxx'] / subdata['r'] ** 3
+    r =  subdata['r'] / AU_TO_ANG
+    Er = (subdata['px'] / AU_TO_DEBYE / r ** 2 + subdata['Qxx'] / AU_TO_DEBYE / AU_TO_ANG / r ** 3) * AU_TO_EV
+    subdata.insert(1, 'Er', Er)
+
+    return subdata
+
+def plot_Er(ax, data: pandas.DataFrame, column: str, family: str, color: str):
+    subdata = data[data['family'] == family]
 
     excluded_ = [int(x.replace('mol_', '')) in EXCLUDE for x in subdata['name']]
     not_excluded_ = [not x for x in excluded_]
 
-    ax.plot(Er[not_excluded_], subdata[not_excluded_][column]  - dG_DH_[not_excluded_], 'o', color=color, label=family.replace('Family.', ''))
-    ax.plot(Er[excluded_], subdata[excluded_][column]  - dG_DH_[excluded_], '^', color=color)
+    ax.plot(subdata[not_excluded_]['Er'], subdata[not_excluded_][column], 'o', color=color, label=family.replace('Family.', ''))
+    ax.plot(subdata[excluded_]['Er'], subdata[excluded_][column], '^', color=color)
 
-    for name, er, e in zip(subdata['name'], Er, subdata[column]  - dG_DH_):
+    for name, er, e in zip(subdata['name'], subdata['Er'], subdata[column]):
         name = name.replace('mol_', '')
         LABELS[column].append(name)
         POINTS_POSITION[column].append((er, e))
         LABELS_KWARGS[column].append(dict(color=color, ha='center', va='center'))
 
 def plot_corr_Er(ax, data: pandas.DataFrame, column: str):
-    subdata = data[(data['solvent'] == 'water') & data['px'].notnull()]
-    subdata = subdata[[int(x.replace('mol_', '')) not in EXCLUDE for x in subdata['name']]]
-
-    if column == 'E_ox':
-        dG_DH_ = dG_DH(subdata['z'] + 1, subdata['z'], subdata['r_ox'] / AU_TO_ANG, subdata['r_rad']  / AU_TO_ANG, 80, 0) * AU_TO_EV
-    else:
-        dG_DH_ = dG_DH(subdata['z'], subdata['z'] - 1, subdata['r_rad'] / AU_TO_ANG,  subdata['r_red']  / AU_TO_ANG, 80, 0) * AU_TO_EV
-
-    result = scipy.stats.linregress(subdata['px'] / subdata['r'] ** 2 + subdata['Qxx'] / subdata['r'] ** 3, subdata[column]  - dG_DH_)
+    subdata = data[[int(x.replace('mol_', '')) not in EXCLUDE for x in data['name']]]
+    result = scipy.stats.linregress(subdata['Er'], subdata[column])
 
-    x = numpy.array([-.2, 1.5])
+    x = numpy.array([-.2, 4])
     ax.plot(x, result.slope*x + result.intercept, 'k--')
-    ax.text(1.2,  1.2*result.slope+result.intercept+.05, '$R^2$={:.2f}'.format(result.rvalue **2))
+    ax.text(3.,  3.*result.slope+result.intercept+.05, '$R^2$={:.2f}'.format(result.rvalue **2))
 
 def make_table(f, data: pandas.DataFrame, solvent: str):
     subdata = data[(data['solvent'] == solvent) & data['px'].notnull()]
@@ -83,51 +84,52 @@ def make_table(f, data: pandas.DataFrame, solvent: str):
 args = parser.parse_args()
 
 data = pandas.read_csv(args.input)
+data = prepare_data(data)
 
 figure = plt.figure(figsize=(7, 10))
 ax3, ax4 = figure.subplots(2, 1)
 
-plot_Er(ax3, data, 'E_ox', 'Family.P6O', 'tab:blue')
-plot_Er(ax3, data, 'E_ox', 'Family.P5O', 'black')
-plot_Er(ax3, data, 'E_ox', 'Family.IIO', 'tab:green')
-plot_Er(ax3, data, 'E_ox', 'Family.APO', 'tab:red')
-plot_corr_Er(ax3, data, 'E_ox')
+plot_Er(ax3, data, 'Ef_ox', 'Family.P6O', 'tab:blue')
+plot_Er(ax3, data, 'Ef_ox', 'Family.P5O', 'black')
+plot_Er(ax3, data, 'Ef_ox', 'Family.IIO', 'tab:green')
+plot_Er(ax3, data, 'Ef_ox', 'Family.APO', 'tab:red')
+plot_corr_Er(ax3, data, 'Ef_ox')
 
 ax3.legend()
 
 positioner = LabelPositioner.from_file(
-    LABELS_PATH['E_ox'], 
-    numpy.array(POINTS_POSITION['E_ox']), 
-    LABELS['E_ox'], 
-    labels_kwargs=LABELS_KWARGS['E_ox']
+    LABELS_PATH['Ef_ox'], 
+    numpy.array(POINTS_POSITION['Ef_ox']), 
+    LABELS['Ef_ox'], 
+    labels_kwargs=LABELS_KWARGS['Ef_ox']
 )
 
 if args.reposition_labels: 
-    positioner.optimize(dx=1e-3, beta=1e4, krep=1, kspring=10000, c=0.02, b0=0.015, scale=(.2, 1))
-    positioner.save(LABELS_PATH['E_ox'])
+    positioner.optimize(dx=1e-3, beta=1e4, krep=1, kspring=10000, c=0.02, b0=0.015, scale=(.1, 1))
+    positioner.save(LABELS_PATH['Ef_ox'])
 
 positioner.add_labels(ax3)
 
-plot_Er(ax4, data, 'E_red', 'Family.P6O', 'tab:blue')
-plot_Er(ax4, data, 'E_red', 'Family.P5O', 'black')
-plot_Er(ax4, data, 'E_red', 'Family.IIO', 'tab:green')
-plot_Er(ax4, data, 'E_red', 'Family.APO', 'tab:red')
-plot_corr_Er(ax4, data, 'E_red')
+plot_Er(ax4, data, 'Ef_red', 'Family.P6O', 'tab:blue')
+plot_Er(ax4, data, 'Ef_red', 'Family.P5O', 'black')
+plot_Er(ax4, data, 'Ef_red', 'Family.IIO', 'tab:green')
+plot_Er(ax4, data, 'Ef_red', 'Family.APO', 'tab:red')
+plot_corr_Er(ax4, data, 'Ef_red')
 
 positioner = LabelPositioner.from_file(
-    LABELS_PATH['E_red'], 
-    numpy.array(POINTS_POSITION['E_red']), 
-    LABELS['E_red'], 
-    labels_kwargs=LABELS_KWARGS['E_red']
+    LABELS_PATH['Ef_red'], 
+    numpy.array(POINTS_POSITION['Ef_red']), 
+    LABELS['Ef_red'], 
+    labels_kwargs=LABELS_KWARGS['Ef_red']
 )
 
 if args.reposition_labels:
-    positioner.optimize(dx=1e-3, beta=1e4,  krep=1, kspring=10000, c=0.02, b0=0.015, scale=(.2, 1))
-    positioner.save(LABELS_PATH['E_red'])
+    positioner.optimize(dx=1e-3, beta=1e4,  krep=1, kspring=10000, c=0.02, b0=0.015, scale=(.1, 1))
+    positioner.save(LABELS_PATH['Ef_red'])
 
 positioner.add_labels(ax4)
 
-[ax.set_xlabel('Electrostatic potential $\\mu_x/r^2 + Q_{xx}/r^3$ (D Å$^{-2}$)') for ax in [ax3, ax4]]
+[ax.set_xlabel('Electrostatic potential $U_q$ (V)') for ax in [ax3, ax4]]
 ax3.set_ylabel('$E^0_{abs}(N^+|N^\\bullet)$ (V)')
 ax4.set_ylabel('$E^0_{abs}(N^\\bullet|N^-)$ (V)')
 

analyses/plot_pot_exp.py

@@ -87,7 +87,7 @@ def plot_corr(ax, data: pandas.DataFrame, solvent: str):
     ax.plot(x, result.slope*x + result.intercept, 'k--')
 
     x = x.min()
-    ax.text(x + .05, result.slope*x + result.intercept, '{:.2f} $\\times E^f_{{rel}}$ + {:.2f}\n($R^2$={:.2f}, MAE={:.2f} V)'.format(result.slope, result.intercept,result.rvalue **2, mae))
+    ax.text(x + .05, result.slope*x + result.intercept, '{:.2f} $\\times E^f_{{rel}}$ + {:.2f} V\n($R^2$={:.2f}, MAE={:.2f} V)'.format(result.slope, result.intercept,result.rvalue **2, mae))
 
 parser = argparse.ArgumentParser()
 parser.add_argument('-i', '--input', default='../data/Data_pot.csv')
@@ -110,9 +110,14 @@ def plot_corr(ax, data: pandas.DataFrame, solvent: str):
 
 plot_exp_vs_theo(ax1, subdata_wa, 'water', 'Family.P6O', 'tab:blue')
 plot_exp_vs_theo(ax1, subdata_wa, 'water', 'Family.P5O', 'black')
+plot_exp_vs_theo(ax1, subdata_wa, 'water', 'Family.IIO', 'tab:green')
+plot_exp_vs_theo(ax1, subdata_wa, 'water', 'Family.APO', 'tab:red')
 
 plot_corr(ax1, subdata_wa, 'water')
 
+ax1.legend()
+ax1.text(0.85, 0.95, 'Water', fontsize=18)
+
 positioner = LabelPositioner.from_file(
     LABELS_PATH['water'], 
     numpy.array(POINTS_POSITION['water']), 
@@ -134,6 +139,7 @@ def plot_corr(ax, data: pandas.DataFrame, solvent: str):
 plot_exp_vs_theo(ax2, subdata_ac, 'acetonitrile', 'Family.APO', 'tab:red')
 
 plot_corr(ax2, subdata_ac, 'acetonitrile')
+ax2.text(0.5, 1.15, 'Acetonitrile', fontsize=18)
 
 positioner = LabelPositioner.from_file(
     LABELS_PATH['acetonitrile'], 
@@ -147,8 +153,7 @@ def plot_corr(ax, data: pandas.DataFrame, solvent: str):
     positioner.save(LABELS_PATH['acetonitrile'])
 
 positioner.add_labels(ax2)
-
-ax2.legend()
+ax2.xaxis.set_major_formatter('{x:.2f}')
 [ax.set_xlabel('Computed $E^f_{rel}(N^+|N^\\bullet)$ (V)') for ax in (ax1, ax2)]
 [ax.set_ylabel('Experimental $E^0_{rel}(N^+|N^\\bullet)$ (V)') for ax in (ax1, ax2)]
 

analyses/plot_pot_matsui.py

@@ -77,7 +77,7 @@ def plot_corr(ax, data: pandas.DataFrame, solvent: str):
     ax.plot(x, result.slope*x + result.intercept, 'k--')
 
     x = x.min()
-    ax.text(x + .05, result.slope*x + result.intercept, '{:.2f} $\\times E^P_{{rel}}$ + {:.2f}\n($R^2$={:.2f}, MAE={:.2f} V)'.format(result.slope, result.intercept, result.rvalue **2, mae))
+    ax.text(x + .05, result.slope*x + result.intercept, '{:.2f} $\\times E^{{Mat}}_{{rel}}$ + {:.2f} V\n($R^2$={:.2f}, MAE={:.2f} V)'.format(result.slope, result.intercept, result.rvalue **2, mae))
 
 parser = argparse.ArgumentParser()
 parser.add_argument('-i', '--input', default='../data/Data_pot.csv')
@@ -101,6 +101,7 @@ def plot_corr(ax, data: pandas.DataFrame, solvent: str):
 plot_exp_vs_matsui(ax1, subdata_wa, 'water', 'Family.P5O', 'black')
 
 plot_corr(ax1, subdata_wa, 'water')
+ax1.text(0.75, 0.9, 'Water', fontsize=18)
 
 positioner = LabelPositioner.from_file(
     LABELS_PATH['water'], 
@@ -125,6 +126,7 @@ def plot_corr(ax, data: pandas.DataFrame, solvent: str):
 plot_exp_vs_matsui(ax2, subdata_ac, 'acetonitrile', 'Family.APO', 'tab:red')
 
 plot_corr(ax2, subdata_ac, 'acetonitrile')
+ax2.text(0.8, 1.15, 'Acetonitrile', fontsize=18)
 
 positioner = LabelPositioner.from_file(
     LABELS_PATH['acetonitrile'], 
@@ -140,6 +142,7 @@ def plot_corr(ax, data: pandas.DataFrame, solvent: str):
 positioner.add_labels(ax2)
 
 ax2.legend()
+ax2.xaxis.set_major_formatter('{x:.2f}')
 [ax.set_xlabel('Computed $E^P_{rel}(N^+|N^\\bullet)$ (V)') for ax in (ax1, ax2)]
 [ax.set_ylabel('Experimental $E^0_{rel}(N^+|N^\\bullet)$ (V)') for ax in (ax1, ax2)]