def createMatrixInteract(self, event):
dlg = RegressDialog(self.parent, "Matrix Interaction Plot")
log = wx.CheckBox(dlg, label="Logistic Fit?")
dlg.Add(log)
if dlg.ShowModal() == wx.ID_OK:
y, xs = dlg.GetValue()
log = log.GetValue()
data = self.parent.data[list(xs) + [y]]
df = data[list(xs)]
fig, axes = plt.subplots(nrows=len(xs), ncols=len(xs))
for i, l1 in enumerate(df):
for j, l2 in enumerate(df):
ax = axes[j, i]
ax.grid(False)
plt.subplot(ax)
if i == j:
sns.regplot(data[l1], data[y], ax=ax),
# would like to do logistic plot, but takes too long
elif i < j:
sns.interactplot(l1, l2, y, data, ax=ax, logistic=log,
cmap=settings["cmap"])
if i != 0 and j != 0:
ax.yaxis.set_visible(False)
if j != len(xs) - 1:
ax.xaxis.set_visible(False)
plt.show()
def make_plot(X_train, y_train, X, y, test_data, model, model_name, features, response):
feature = X.columns
f, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, sharey=False)
sns.regplot(X[feature[4]], y, test_data, ax=ax1)
sns.boxplot(X[feature[4]], y, color="Blues_r", ax=ax2)
model.fit(X_train, y_train)
sns.residplot(X[feature[4]], (model.predict(X) - y) ** 2, color="indianred", lowess=True, ax=ax3)
if model_name is 'linear':
sns.interactplot(X[feature[3]], X[feature[4]], y, ax=ax4, filled=True, scatter_kws={"color": "dimgray"}, contour_kws={"alpha": .5})
elif model_name is 'logistic':
pal = sns.blend_palette(["#4169E1", "#DFAAEF", "#E16941"], as_cmap=True)
levels = np.linspace(0, 1, 11)
sns.interactplot(X[feature[3]], X[feature[4]], y, levels=levels, cmap=pal, logistic=True)
else:
pass
ax1.set_title('Regression')
ax2.set_title(feature[4]+' Value')
ax3.set_title(feature[4]+' Residuals')
ax4.set_title('Two-value Interaction')
f.tight_layout()
plt.savefig(model_name+'_'+feature[4], bbox_inches='tight')
# Multi-variable correlation significance level
f, ax = plt.subplots(figsize=(10, 10))
cmap = sns.blend_palette(["#00008B", "#6A5ACD", "#F0F8FF",
"#FFE6F8", "#C71585", "#8B0000"], as_cmap=True)
sns.corrplot(test_data, annot=False, diag_names=False, cmap=cmap)
ax.grid(False)
ax.set_title('Multi-variable correlation significance level')
plt.savefig(model_name+'_multi-variable_correlation', bbox_inches='tight')
# complete coefficient plot - believe this is only for linear regression
sns.coefplot("diagnosis ~ "+' + '.join(features), test_data, intercept=True)
plt.xticks(rotation='vertical')
plt.savefig(model_name+'_coefficient_effects', bbox_inches='tight')
def interactplot(ax):
rs = np.random.RandomState(11)
n = 80
x1 = rs.randn(n)
x2 = x1 / 5 + rs.randn(n)
b0, b1, b2, b3 = 1.5, 4, -1, 3
y = b0 + b1 * x1 + b2 * x2 + b3 * x1 * x2 + rs.randn(n)
sns.interactplot(x1, x2, y, colorbar=False, ax=ax)
ax.set_title("interactplot()")
def interactplot(ax):
rs = np.random.RandomState(11)
n = 80
x1 = rs.randn(n)
x2 = x1 / 5 + rs.randn(n)
b0, b1, b2, b3 = 1.5, 4, -1, 3
y = b0 + b1 * x1 + b2 * x2 + b3 * x1 * x2 + rs.randn(n)
sns.interactplot(x1, x2, y, colorbar=False, ax=ax)
ax.set_title("interactplot()")
def _update_plot(self, axis, view, style):
style.pop('zorder', None)
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 == '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, style)
def bvContour(x1, x2, y, **kwargs):
contour_plot = plt.figure()
outfile = kwargs.pop("savefig", None)
sns.interactplot(x1,
x2,
y,
filled=True,
scatter_kws={
"marker": "x",
"markersize": 1
},
contour_kws={"linewidths": 2},
**kwargs)
if outfile:
plt.title(kwargs.pop("title", ""))
contour_plot.savefig(outfile)
plt.close()
return contour_plot
def createInteraction(self, event):
dlg = GraphDialog(self.parent, "Matrix Plot Input", ("X1", "X2", "Y"),
size=(700, 200), add=False, groups=False)
fill = wx.CheckBox(dlg, label="Fill")
fill.SetValue(True)
log = wx.CheckBox(dlg, label="Logistic Fit?")
dlg.Add(fill)
dlg.Add(log)
if dlg.ShowModal() == wx.ID_OK:
(x1, x2, y), fill = dlg.GetName()[0], fill.GetValue()
log = log.GetValue()
data = self.parent.data[[x1, x2, y]].astype(float)
dlg.Destroy()
temp = data[[x1, x2, y]]
sns.interactplot(x1, x2, y, temp, cmap=settings["cmap"], filled=fill,
logistic=log)
plt.show()
def interact(model, data, x, y, z, **kwargs):
df = data.as_dataframe(_combine_keys(x, y, z))
sns.interactplot(x, y, z, df, **kwargs)
parameters["density"] = mu["density"]
parameters["energy"] = mu["energy"]
parameters["temperature"] = temperature
data.append(parameters)
data = pd.DataFrame(data).dropna()
Q = data.pivot_table(index=["q0", "sigma0"], columns=["temperature"], values=["energy", "density"])
Q = data.pivot_table(index=["q0", "sigma0"], columns="temperature", values="energy")
dE0 = (Q[300.] - Q[280.])
dE0.name = "dE0"
dE0 = dE0.reset_index()
dE0.dropna(inplace=True)
temperature = 280.
ind = data.temperature == temperature
mydata = data[ind]
figure()
sns.interactplot("q0", "sigma0", "density", mydata, cmap="coolwarm", filled=True, levels=25);
figure()
sns.interactplot("q0", "sigma0", "dE0", dE0, cmap="coolwarm", filled=True)
data["sigmasum"] = data["sigma0"] + data["sigma1"]
data["epsilonsqrt"] = (data["epsilon0"] * data["epsilon1"]) ** 0.5
figure()
sns.interactplot("sigmasum", "epsilonsqrt", "density", data, cmap="coolwarm", filled=True)
"""
Continuous interactions
=======================
"""
import numpy as np
import pandas as pd
import seaborn as sns
sns.set(style="darkgrid")
# Generate a random dataset with strong simple effects and an interaction
n = 80
rs = np.random.RandomState(11)
x1 = rs.randn(n)
x2 = x1 / 5 + rs.randn(n)
b0, b1, b2, b3 = .5, .25, -1, 2
y = b0 + b1 * x1 + b2 * x2 + b3 * x1 * x2 + rs.randn(n)
df = pd.DataFrame(np.c_[x1, x2, y], columns=["x1", "x2", "y"])
# Show a scatterplot of the predictors with the estimated model surface
sns.interactplot("x1", "x2", "y", df)
#On the diagonals we have bivariate kernel density estimation
g.map_offdiag(sns.kdeplot, cmap="Blues_d", n_levels=6);
#It is often useful to give some relationships a closer look
sns.set()
g = sns.jointplot("educ", "loyalty", data=reduced, kind="reg", color="b")
#Violin plots are a good way to not learn about distributions
sns.set_style("whitegrid")
sns.violinplot(y='loyalty', data=reduced)
#also distributions that are affected by independent variables
sns.violinplot(x='age', y='loyalty', hue='gender', data=reduced, split=True)
#Interaction plots are a great way to learn about multivariate interactions
sns.interactplot(x1='Emotional', x2='Trust', y='loyalty', data=reduced)
#Step 3
#Think! Come up with some hypothesis!
#Step 4 Test
#the usual social science tests
#Hypothesis: Emotional people are more loyal
#Simple linear regression
res = sm.ols(formula="loyalty ~ Emotional", data=reduced).fit()
res.summary()
#Answer: Yes they are but that is not enought to explain loyalty
#data = pd.read_hdf("./symmetric-grid2.h5", 'data')
data = pd.read_hdf("./ccl4-grid.h5", 'data')
Q = data.pivot_table(index=keys,
columns=["temperature"],
values=["energy", "density", "dielectric", "kappa"])
temperature = 280.
ind = data.temperature == temperature
mydata = data[ind]
figure()
sns.interactplot(keys[0],
keys[1],
"density",
mydata,
cmap="coolwarm",
filled=True,
levels=25)
figure()
sns.interactplot(keys[0],
keys[1],
"evap",
mydata,
cmap="coolwarm",
filled=True,
levels=25)
figure()
sns.interactplot(keys[0],
# axarr[1, 1].set_title('Axis [1,1]')
# In[166]:
X.cholesterol
# In[167]:
from sklearn.grid_search import GridSearchCV
# In[168]:
sns.interactplot(X[X.columns[4]], X[X.columns[5]], y.values, ax=ax4, cmap="coolwarm", filled=True, levels=25)
# In[169]:
y.values
# In[170]:
from sklearn.svm import SVC
linear_reg = LinearRegression()
parameters = {'normalize':(True, False)}
# In[306]:
import seaborn as sns
sns.set()
from astropy.table import Table
import pandas
import numpy as np
import sys
# table = Table.read(sys.argv[1])
table = pandas.read_csv(sys.argv[1]).dropna(how='all')
# table.pivot("AZ", "ALT")
ax1 = sns.interactplot("AZ", "ALT", "AIRMASS", data=table)
sns.plt.show()
def interact(model, data, x, y, z, **kwargs):
df = data.as_dataframe(_combine_keys(x, y, z))
sns.interactplot(x, y, z, df, **kwargs)
"""
Continuous interactions
=======================
"""
import numpy as np
import pandas as pd
import seaborn as sns
sns.set(style="darkgrid")
# Generate a random dataset with strong simple effects and an interaction
n = 80
rs = np.random.RandomState(11)
x1 = rs.randn(n)
x2 = x1 / 5 + rs.randn(n)
b0, b1, b2, b3 = .5, .25, -1, 2
y = b0 + b1 * x1 + b2 * x2 + b3 * x1 * x2 + rs.randn(n)
df = pd.DataFrame(np.c_[x1, x2, y], columns=["x1", "x2", "y"])
# Show a scatterplot of the predictors with the estimated model surface
sns.interactplot("x1", "x2", "y", df)
# ----------------------
sns.set(style="darkgrid")
ax = plt.subplot(gs[2:4, 1])
plt.title("interactplot()")
rs = np.random.RandomState(11)
n = 80
x1 = rs.randn(n)
x2 = x1 / 5 + rs.randn(n)
b0, b1, b2, b3 = 1.5, 4, -1, 3
y = b0 + b1 * x1 + b2 * x2 + b3 * x1 * x2 + rs.randn(n)
sns.interactplot(x1, x2, y, colorbar=False, ax=ax)
# Correlation matrix
# ------------------
ax = plt.subplot(gs[4:, 0])
rs = np.random.RandomState(0)
x0, x1 = rs.randn(2, 60)
x2, x3 = rs.multivariate_normal([0, 0], [(1, -.5), (-.5, 1)], 60).T
x2 += x0 / 8
x4 = x1 + rs.randn(60) * 2
data = np.c_[x0, x1, x2, x3, x4]
sns.corrplot(data, ax=ax)
import pymc
import sklearn.gaussian_process
import os
import pandas as pd
import glob
#keys = ["q0", "sigma0"]
keys = ["q0", "sigma1"]
#data = pd.read_hdf("./symmetric-grid2.h5", 'data')
data = pd.read_hdf("./ccl4-grid.h5", 'data')
Q = data.pivot_table(index=keys, columns=["temperature"], values=["energy", "density", "dielectric", "kappa"])
temperature = 280.
ind = data.temperature == temperature
mydata = data[ind]
figure()
sns.interactplot(keys[0], keys[1], "density", mydata, cmap="coolwarm", filled=True, levels=25);
figure()
sns.interactplot(keys[0], keys[1], "evap", mydata, cmap="coolwarm", filled=True, levels=25);
figure()
sns.interactplot(keys[0], keys[1], "dielectric", mydata, cmap="coolwarm", filled=True, levels=25);
figure()
sns.interactplot(keys[0], keys[1], "kappa", mydata, cmap="coolwarm", filled=True, levels=25);
