def _update_plot(self, axis, view):
if self.plot_type == 'regplot':
sns.regplot(x=view.x,
y=view.y,
data=view.data,
ax=axis,
**self.style)
elif self.plot_type == 'boxplot':
self.style.pop('return_type', None)
self.style.pop('figsize', None)
sns.boxplot(view.data[view.y],
view.data[view.x],
ax=axis,
**self.style)
elif self.plot_type == 'violinplot':
sns.violinplot(view.data[view.y],
view.data[view.x],
ax=axis,
**self.style)
elif self.plot_type == 'interact':
sns.interactplot(view.x,
view.x2,
view.y,
data=view.data,
ax=axis,
**self.style)
elif self.plot_type == 'corrplot':
sns.corrplot(view.data, ax=axis, **self.style)
elif self.plot_type == 'lmplot':
sns.lmplot(x=view.x,
y=view.y,
data=view.data,
ax=axis,
**self.style)
elif self.plot_type in ['pairplot', 'pairgrid', 'facetgrid']:
map_opts = [(k, self.style.pop(k)) for k in self.style.keys()
if 'map' in k]
if self.plot_type == 'pairplot':
g = sns.pairplot(view.data, **self.style)
elif self.plot_type == 'pairgrid':
g = sns.PairGrid(view.data, **self.style)
elif self.plot_type == 'facetgrid':
g = sns.FacetGrid(view.data, **self.style)
for opt, args in map_opts:
plot_fn = getattr(sns, args[0]) if hasattr(
sns, args[0]) else getattr(plt, args[0])
getattr(g, opt)(plot_fn, *args[1:])
plt.close(self.handles['fig'])
self.handles['fig'] = plt.gcf()
else:
super(SNSFramePlot, self)._update_plot(axis, view)
def _update_plot(self, axis, view):
style = self._process_style(self.style[self.cyclic_index])
if self.plot_type == 'factorplot':
opts = dict(style, **({'hue': view.x2} if view.x2 else {}))
sns.factorplot(x=view.x, y=view.y, data=view.data, **opts)
elif self.plot_type == 'regplot':
sns.regplot(x=view.x, y=view.y, data=view.data, ax=axis, **style)
elif self.plot_type == 'boxplot':
style.pop('return_type', None)
style.pop('figsize', None)
sns.boxplot(view.data[view.y], view.data[view.x], ax=axis, **style)
elif self.plot_type == 'violinplot':
if view.x:
sns.violinplot(view.data[view.y],
view.data[view.x],
ax=axis,
**style)
else:
sns.violinplot(view.data, ax=axis, **style)
elif self.plot_type == 'interact':
sns.interactplot(view.x,
view.x2,
view.y,
data=view.data,
ax=axis,
**style)
elif self.plot_type == 'corrplot':
sns.corrplot(view.data, ax=axis, **style)
elif self.plot_type == 'lmplot':
sns.lmplot(x=view.x, y=view.y, data=view.data, ax=axis, **style)
elif self.plot_type in ['pairplot', 'pairgrid', 'facetgrid']:
style_keys = list(style.keys())
map_opts = [(k, style.pop(k)) for k in style_keys if 'map' in k]
if self.plot_type == 'pairplot':
g = sns.pairplot(view.data, **style)
elif self.plot_type == 'pairgrid':
g = sns.PairGrid(view.data, **style)
elif self.plot_type == 'facetgrid':
g = sns.FacetGrid(view.data, **style)
for opt, args in map_opts:
plot_fn = getattr(sns, args[0]) if hasattr(
sns, args[0]) else getattr(plt, args[0])
getattr(g, opt)(plot_fn, *args[1:])
plt.close(self.handles['fig'])
self.handles['fig'] = plt.gcf()
else:
super(SNSFramePlot, self)._update_plot(axis, view)
def _update_plot(self, axis, view):
style = self._process_style(self.style[self.cyclic_index])
if self.plot_type == 'factorplot':
opts = dict(style, **({'hue': view.x2} if view.x2 else {}))
sns.factorplot(x=view.x, y=view.y, data=view.data, **opts)
elif self.plot_type == 'regplot':
sns.regplot(x=view.x, y=view.y, data=view.data,
ax=axis, **style)
elif self.plot_type == 'boxplot':
style.pop('return_type', None)
style.pop('figsize', None)
sns.boxplot(view.data[view.y], view.data[view.x], ax=axis,
**style)
elif self.plot_type == 'violinplot':
if view.x:
sns.violinplot(view.data[view.y], view.data[view.x], ax=axis,
**style)
else:
sns.violinplot(view.data, ax=axis, **style)
elif self.plot_type == 'interact':
sns.interactplot(view.x, view.x2, view.y,
data=view.data, ax=axis, **style)
elif self.plot_type == 'corrplot':
sns.corrplot(view.data, ax=axis, **style)
elif self.plot_type == 'lmplot':
sns.lmplot(x=view.x, y=view.y, data=view.data,
ax=axis, **style)
elif self.plot_type in ['pairplot', 'pairgrid', 'facetgrid']:
style_keys = list(style.keys())
map_opts = [(k, style.pop(k)) for k in style_keys if 'map' in k]
if self.plot_type == 'pairplot':
g = sns.pairplot(view.data, **style)
elif self.plot_type == 'pairgrid':
g = sns.PairGrid(view.data, **style)
elif self.plot_type == 'facetgrid':
g = sns.FacetGrid(view.data, **style)
for opt, args in map_opts:
plot_fn = getattr(sns, args[0]) if hasattr(sns, args[0]) else getattr(plt, args[0])
getattr(g, opt)(plot_fn, *args[1:])
if self._close_figures:
plt.close(self.handles['fig'])
self.handles['fig'] = plt.gcf()
else:
super(SNSFramePlot, self)._update_plot(axis, view)
def plot_synapse_2d(X, zh, output='synapse_cluster_2d.pdf'):
pca = PCA(n_components=2)
X_new = pca.fit_transform(X)
d = {
#r'$x_1$': X_new[:,0]/(10**5),
r'$x_1$': X_new[:,0],
#r'$x_2$': X_new[:,1]/(10**6),
r'$x_2$': X_new[:,1],
r'$\mathcal{C}_j$': [int(z+1) for z in zh]
}
df = pd.DataFrame(data=d)
g = sns.lmplot(r'$x_1$', r'$x_2$',
data=df, hue=r'$\mathcal{C}_j$',
fit_reg=False, scatter=True, scatter_kws={"s":4})
g.set(xlabel=r'$x_1$')
g.set(ylabel=r'$x_2$')
#g.set(xlabel=r'$x_1$~$(\times 10^6)$')
#g.set(ylabel=r'$x_2$~$(\times 10^5)$')
g.savefig(output)
def plot_synapse_2d(X, zh, output='synapse_cluster_2d.pdf'):
pca = PCA(n_components=2)
X_new = pca.fit_transform(X)
d = {
#r'$x_1$': X_new[:,0]/(10**5),
r'$x_1$': X_new[:, 0],
#r'$x_2$': X_new[:,1]/(10**6),
r'$x_2$': X_new[:, 1],
r'$\mathcal{C}_j$': [int(z + 1) for z in zh]
}
df = pd.DataFrame(data=d)
g = sns.lmplot(r'$x_1$',
r'$x_2$',
data=df,
hue=r'$\mathcal{C}_j$',
fit_reg=False,
scatter=True,
scatter_kws={"s": 4})
g.set(xlabel=r'$x_1$')
g.set(ylabel=r'$x_2$')
#g.set(xlabel=r'$x_1$~$(\times 10^6)$')
#g.set(ylabel=r'$x_2$~$(\times 10^5)$')
g.savefig(output)
def lmplot(self, x=None, y=None, hue=None, data=None, *args, **kwargs):
"""
Plot data and regression model fits
Parameters
----------
x : a list of names of variable in data that need to visualize \
their distribution
y : a list of names of variable in data that need to visualize \
its joint distribution against every x above
hue : the name of a variable in data that provides labels for each \
category
data : pandas dataframe
**kwargs : other arguments in seaborn.jointplot
palette : palette name, list, or dict, optional
col_wrap : int, optional
size : scalar, optional
aspect : scalar, optional
markers : matplotlib marker code or list of marker codes, optional
share{x,y} : bool, optional
legend : bool, optional
legend_out : bool, optional
x_estimator : callable that maps vector -> scalar, optional
x_bins : int or vector, optional
x_ci : 'ci', 'sd', int in [0, 100] or None, optional
scatter : bool, optional
fit_reg : bool, optional
ci : int in [0, 100] or None, optional
n_boot : int, optional
units : variable name in data, optional
order : int, optional
logistic : bool, optional
lowess : bool, optional
robust : bool, optional
logx : bool, optional
{x,y}_partial : strings in data or matrices
truncate : bool, optional
{x,y}_jitter : floats, optional
{scatter,line}_kws : dictionaries
Returns
-------
JointGrid object with the plot on it
References
----------
Seaborn lmplot further documentation
https://seaborn.pydata.org/generated/seaborn.seaborn.lmplot
"""
# check data
if not isinstance(data, (pd.DataFrame)):
raise ValueError('data must be pandas dataframe')
# check x and y
if x is None:
raise ValueError('x can NOT be None')
else: # x is NOT None
if not isinstance(x, (list, tuple, np.ndarray, pd.Index)):
x = [x]
if y is None:
raise ValueError('y can NOT be None')
else: # y is NOT None
if not isinstance(y, (list, tuple, np.ndarray, pd.Index)):
y = [y]
if hue is not None:
if hue not in data.columns.values:
raise ValueError('{} is NOT in data'.format(hue))
# no figure configuration needed
plt.close()
# iterate thru x
for i, col_y in enumerate(y):
if col_y not in data.columns.values:
raise ValueError('{} is NOT in data'.format(col_y))
b = data[col_y]
b_not_nan = np.ones(b.shape[0], dtype=np.bool)
if np.logical_not(np.isfinite(b)).any():
logger.warning('RUNTIME WARNING: {} column has inf or nan '
''.format(col_y))
b = b.replace([-np.inf, np.inf], np.nan)
# filter
b_not_nan = np.logical_not(b.isnull())
for j, col_x in enumerate(x):
# check if col in data
if col_x not in data.columns.values:
raise ValueError('{} is NOT in data'.format(col_x))
a = data[col_x]
a_not_nan = np.ones(a.shape[0], dtype=np.bool)
if np.logical_not(np.isfinite(a)).any():
logger.warning('RUNTIME WARNING: {} column has inf or '
'nan'.format(col_x))
a = a.replace([-np.inf, np.inf], np.nan)
# filter
a_not_nan = np.logical_not(a.isnull())
# joint filter
not_nan = b_not_nan & a_not_nan
joint_grid = sns.lmplot(
x=col_x,
y=col_y,
data=data.loc[not_nan, :],
hue=hue,
legend=True,
legend_out=False,
# size=self.size[0],
*args,
**kwargs)
joint_grid.fig.axes[0].set_title(label='Reg Fit of {} on {} '
''.format(col_y, col_x),
fontsize=self.title_fontsize)
joint_grid.fig.axes[0].set_xlabel(xlabel=col_x,
fontsize=self.label_fontsize)
joint_grid.fig.axes[0].set_ylabel(ylabel=col_y,
fontsize=self.label_fontsize)
joint_grid.fig.axes[0].tick_params(
axis='both', which='maj', labelsize=self.tick_fontsize)
joint_grid.fig.axes[0].legend(loc='upper right')
joint_grid.fig.subplots_adjust(wspace=0.5,
hspace=0.3,
left=0.125,
right=0.9,
top=0.9,
bottom=0.1)
joint_grid.fig.tight_layout()
plt.show()
# With red crosses.
plt.figure(2)
plt.plot(CarData['hwy'], CarData['cty'], 'r+')
plt.xlabel('hwy')
plt.ylabel('cty')
plt.title('Scatter plot of hwy vs cty')
plt.grid(True)
plt.show()
# You can also change the size of the points depending on a variable
# E.g. if you want to display more common brands in terms of observations
# in a larger way, use "size":
CarData['counts'] = CarData.groupby(['make'])['make'].transform('count')
plt.figure(3)
plt.scatter(CarData['hwy'],
CarData['cty'],
marker='o',
c='r',
s=CarData['counts'])
plt.xlabel('hwy')
plt.ylabel('cty')
plt.title('Scatter plot of hwy vs cty')
plt.grid(True)
plt.show()
# We use sns.regplot or sns.lmplot also
sns.regplot(x='cty', y='hwy', marker="+", ci=95, data=CarData)
sns.lmplot(x="cty", y="hwy", marker="o", ci=95, data=CarData)
plt.pie(peace_age, labels=peace_age.index, autopct='%1.1f%%')
plt.show()
sns.jointplot(x="Year",
y="Age",
kind='reg',
data=data)
plt.show()
sns.boxplot(data=data,
x='Category',
y='Age')
plt.show()
sns.lmplot('Year','Age',data=data,lowess=True, aspect=2, line_kws={'color' : 'black'})
plt.show()
# Question 2: What words are most frequently written in the prize motivation?
top_N = 10
stopwords = nltk.corpus.stopwords.words('english')
re_stopwords = r'\b(?:{})\b'.format('|'.join(stopwords))
words = (data['Motivation']
.str.lower()
.replace([r'\|', re_stopwords], [' ', ' '], regex=True)
.str.cat(sep=' ')
.split()
)
def plot(self, show_samples, show_loadings, sbrn_plt):
# Normalizer and Delta perform badly
# They flatten out all difference in a PCA plot
pca = PCA(n_components=self.n_components)
X_bar = pca.fit_transform(self.X)
var_exp = pca.explained_variance_ratio_
var_pc1 = np.round(var_exp[0]*100, decimals=2)
var_pc2 = np.round(var_exp[1]*100, decimals=2)
explained_variance = np.round(sum(pca.explained_variance_ratio_)*100, decimals=2)
comps = pca.components_
comps = comps.transpose()
loadings = pca.components_.transpose()
vocab_weights_p1 = sorted(zip(self.features, comps[:,0]), key=lambda tup: tup[1], reverse=True)
vocab_weights_p2 = sorted(zip(self.features, comps[:,1]), key=lambda tup: tup[1], reverse=True)
if sbrn_plt == False:
# Generate color dictionary
color_dict = {author:index for index, author in enumerate(sorted(set(self.authors)))}
cmap = discrete_cmap(len(color_dict), base_cmap='brg')
if show_samples == True:
fig = plt.figure(figsize=(8,6))
ax = fig.add_subplot(111)
x1, x2 = X_bar[:,0], X_bar[:,1]
# If anything needs to be invisible in plot, add to exclusion_list
ax.scatter(x1, x2, 100, edgecolors='none', facecolors='none', cmap='rainbow')
for index, (p1, p2, a, title) in enumerate(zip(x1, x2, self.authors, self.titles)):
ax.scatter(p1, p2, marker='o', color=cmap(color_dict[a]), s=20)
ax.text(p1, p2, title.split('_')[-1], color='black', fontdict={'size': 5})
# Legend settings (code for making a legend)
collected_patches = []
for author in set(self.authors):
legend_patch = mpatches.Patch(color=cmap(color_dict[author]), label=author.split('-')[0])
collected_patches.append(legend_patch)
plt.legend(handles=collected_patches, fontsize=7)
ax.set_xlabel('Principal Component 1 \n \n Explained Variance: {}% \n Sample Size: {} words/sample \n Number of Features: {} features'.format(str(explained_variance), str(self.sample_size), str(len(self.features))), fontdict={'size': 7})
ax.set_ylabel('Principal Component 2', fontdict={'size': 7})
if show_loadings == True:
ax2 = ax.twinx().twiny()
l1, l2 = loadings[:,0], loadings[:,1]
ax2.scatter(l1, l2, 100, edgecolors='none', facecolors='none');
for x, y, l in zip(l1, l2, self.features):
ax2.text(x, y, l, ha='center', va="center", color="black",
fontdict={'family': 'Arial', 'size': 6})
# Align axes
# Important to adjust margins first when function words fall outside plot
# This is due to the axes aligning (def align).
ax2.margins(x=0.14, y=0.14)
align_xaxis(ax, 0, ax2, 0)
align_yaxis(ax, 0, ax2, 0)
plt.axhline(y=0, ls="--", lw=0.5, c='0.75')
plt.axvline(x=0, ls="--", lw=0.5, c='0.75')
plt.tight_layout()
plt.show()
elif show_loadings == False:
plt.axhline(y=0, ls="--", lw=0.5, c='0.75')
plt.axvline(x=0, ls="--", lw=0.5, c='0.75')
plt.tight_layout()
plt.show()
# Converting PDF to PNG, use pdftoppm in terminal and -rx -ry for resolution settings
fig.savefig(os.path.dirname(os.getcwd()) + "/pca.pdf", transparent=True, format='pdf')
elif show_samples == False:
fig = plt.figure(figsize=(8, 6))
ax2 = fig.add_subplot(111)
l1, l2 = loadings[:,0], loadings[:,1]
ax2.scatter(l1, l2, 100, edgecolors='none', facecolors='none')
for x, y, l in zip(l1, l2, features):
ax2.text(x, y, l, ha='center', va='center', color='black',
fontdict={'family': 'Arial', 'size': 6})
ax2.set_xlabel('PC1')
ax2.set_ylabel('PC2')
align_xaxis(ax, 0, ax2, 0)
align_yaxis(ax, 0, ax2, 0)
plt.axhline(y=0, ls="--", lw=0.5, c='0.75')
plt.axvline(x=0, ls="--", lw=0.5, c='0.75')
plt.tight_layout()
plt.show()
fig.savefig(os.path.dirname(os.getcwd()) + "/pca.pdf", bbox_inches='tight', transparent=True, format='pdf')
# Converting PDF to PNG, use pdftoppm in terminal and -rx -ry for resolution settings
else:
data = [(title.split("_")[0], author, pc1, pc2) for [pc1, pc2], title, author in zip(X_bar, self.titles, self.authors)]
df = pd.DataFrame(data, columns=['title', 'author', 'PC1', 'PC2'])
# Get the x in an array
sns.set_style('darkgrid')
sns_plot = sns.lmplot('PC1', 'PC2', data=df, fit_reg=False, hue="author",
scatter_kws={"marker": "+","s": 100}, markers='o', legend=False)
plt.legend(loc='upper right')
plt.tight_layout()
plt.show()
sns_plot.savefig(os.path.dirname(os.getcwd()) + "/pca.pdf")
