I have a difficulty in the customization of the pairplot.
1) The kde plots in the diagonal are not colored by Class
2) The plots in the diagonal do not fit and get cropped
3) I would like to control the font size of the title of the legend
Finally a get a message I do not understand:
C:\ProgramData\Anaconda3\lib\site-packages\statsmodels\nonparametric\kde.py:494: RuntimeWarning: invalid value encountered in true_divide
binned = fast_linbin(X,a,b,gridsize)/(delta*nobs)
C:\ProgramData\Anaconda3\lib\site-packages\statsmodels\nonparametric\kdetools.py:34: RuntimeWarning: invalid value encountered in double_scalars
FAC1 = 2*(np.pi*bw/RANGE)**2
My code using a reproducible example is here:
import pandas as pd
import numpy as np
import matplotlib
from matplotlib import pyplot as plt
import seaborn as sns
%matplotlib inline
from sklearn import datasets
iris = datasets.load_iris()
df = np.concatenate( (iris.data, np.matrix(iris.target).T), axis = 1)
df1 = pd.DataFrame(df, columns = iris.feature_names + ['Class'])
SMALL_SIZE = 20
MEDIUM_SIZE = 25
BIGGER_SIZE = 30
plt.rc('font', size=SMALL_SIZE) # controls default text sizes
plt.rc('axes', titlesize=SMALL_SIZE) # fontsize of the axes title
plt.rc('axes', labelsize=MEDIUM_SIZE) # fontsize of the x and y labels
plt.rc('xtick', labelsize=SMALL_SIZE) # fontsize of the tick labels
plt.rc('ytick', labelsize=SMALL_SIZE) # fontsize of the tick labels
plt.rc('legend', fontsize=MEDIUM_SIZE) # legend fontsize
plt.rc('figure', titlesize=BIGGER_SIZE) # fontsize of the figure title
sns.pairplot(df1, hue = 'Class', diag_kind = 'kde', plot_kws = {'alpha': 0.6, 's': 80, 'edgecolor': 'k'}, size = 6);
To solve 1) and 2) update your seaborn to version 0.8.1. Possibly update matplotlib as well.
To solve 3) assign the pairplot to a variable g and call
g._legend.get_title().set_fontsize(20)
For the warning you get, this is due to the "Class" column being part of the grid. This does not make too much sense anyways, so leave that one out by specifying the variables to grid, in this case vars = iris.feature_names,.
Complete code:
import pandas as pd
import numpy as np
from matplotlib import pyplot as plt
import seaborn as sns
from sklearn import datasets
iris = datasets.load_iris()
df = np.concatenate( (iris.data, np.matrix(iris.target).T), axis = 1)
df1 = pd.DataFrame(df, columns = iris.feature_names + ['Class'])
g = sns.pairplot(df1, vars = iris.feature_names, hue = 'Class', diag_kind = 'kde',
plot_kws = {'alpha': 0.6, 's': 80, 'edgecolor': 'k'}, size = 2);
g._legend.get_title().set_fontsize(20)
plt.show()
Related
I want to divide and color points,val_lab(611,3) by their labels,pred_LAB(611,)
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
fig = plt.figure()
ax = plt.axes(projection = '3d')
ax.set_xlabel('L')
ax.set_ylabel('A')
ax.set_zlabel('B')
for i in range(0, len(val_lab)):
ax.scatter3D(
val_lab[i,0],
val_lab[i,1],
val_lab[i,2],
s = 8,
marker='o',
c = pred_LAB
#cmap = 'rainbow'
)
#ax.legend(*points.legend_elements(), title = 'clusters')
plt.show()
The problem is it shows error,
c' argument has 611 elements, which is not acceptable for use with 'x'
with size 1, 'y' with size 1.
However, if the dataset only have ten points,it can show the figure correctly, I don't know how to solve this problem, besides, how to add legend of this figure?
In your solution you would want to replace c = pred_LAB with c = pred_LAB[i]. But you do not have to use a for loop to plot the data. You can just use the following:
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
# generate random input data
val_lab = np.random.randint(0,10,(611,3))
pred_LAB = np.random.uniform(0,1, (611,))
# plot data
fig = plt.figure()
ax = plt.axes(projection = '3d')
ax.set_xlabel('L')
ax.set_ylabel('A')
ax.set_zlabel('B')
# create one 3D scatter plot - no for loop
ax.scatter3D(
val_lab[:,0],
val_lab[:,1],
val_lab[:,2],
s = 8,
marker='o',
c = pred_LAB,
cmap = 'rainbow',
label='my points'
)
# add legend
plt.legend()
plt.show()
import pandas as pd
import seaborn as sns
from matplotlib import pyplot as plt
import matplotlib.pyplot as plt
sns.set(style="darkgrid")
df = pd.read_csv('Leap_Static_trials.csv')
Length = sns.swarmplot(x='name', y= 'length', data= df, color = 'green')
Width = sns.swarmplot(x='name', y= 'width', data= df, color = 'red')
plt.legend(labels=['Length','Width'])
plt.show()
From my dataset df I am plotting the length and width of the fingers taken from Leap Motion Controller. I am unable to change the legend to include the second color (red) which signifies the width.
Please find the attached figure as well. Your help is much appreciated. :)
Adding the parameter label= to a plot command usually creates the legend handles and labels automatically. In this case, seaborn creates handles for each column (so 5 of each). A trick is to create the legend with only the first and the last of the handles and the labels.
import pandas as pd
import seaborn as sns
from matplotlib import pyplot as plt
import matplotlib.pyplot as plt
sns.set(style="darkgrid")
N = 100
# df = pd.read_csv('Leap_Static_trials.csv')
names = list('abcde')
ax = plt.gca()
df = pd.DataFrame({'name': np.random.choice(names, N),
'length': np.random.normal(50, 0.7, N),
'width': np.random.normal(20, 0.5, N)})
Length = sns.swarmplot(x='name', y='length', data=df, color='green', label='Length', order=names, ax=ax)
Width = sns.swarmplot(x='name', y='width', data=df, color='red', label='Width', ax=ax)
handles, labels = ax.get_legend_handles_labels()
plt.legend([handles[0], handles[-1]], [labels[0], labels[-1]])
plt.show()
I want to fill the gridded map with colors based on the value of interest. A sample data is here:
import pandas as pd
df = pd.DataFrame()
df['lon'] = [100,105,110,115,120,125,130]
df['lat'] = [38,40,42,44,46,48,50]
df['value'] = [1,2,3,4,5,6,7]
Specifically, is it possible to do this with Cartopy? I found a similar question here:https://stackoverflow.com/questions/53412785/plotting-pandas-csv-data-onto-cartopy-map. But that post was to plot scattered points, I need to fill the grids with colors.
I myself tried:
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import cartopy.crs as ccrs
lon, lat = np.meshgrid(df['lon'], df['lat'])
fig = plt.figure(figsize=[15,15])
ax = plt.axes(projection=ccrs.PlateCarree())
ax.pcolormesh(lon,lat,df['variable'],latlon=True,cmap='jet')
plt.show()
The error is at ax.pcolormesh(...), it says "not enough values to unpack (expected 2, got 1)"
Many thanks for your help.
For discrete data you can create rectangular patches for each point. Here is a possible solution for your sample data set. Each row of data (lat, long, value) is used to create a rectangular patch. The value is normalized by dividing with max(value) to enable using colormap for coloring the patches.
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import cartopy.crs as ccrs
import matplotlib.patches as mpatches
def make_rect(clon, clat, dlon, dlat):
lon_min = clon - dlon/2.
lat_min = clat - dlat/2.
lon_max = clon + dlon/2.
lat_max = clat + dlat/2.
# clockwise from LL
#lons = [lon_min, lon_min, lon_max, lon_max, lon_min]
#lats = [lat_min, lat_max, lat_max, lat_min, lat_min]
ll = [lon_min,lat_min]
ul = [lon_min,lat_max]
ur = [lon_max,lat_max]
lr = [lon_max,lat_min]
return [ll, ul, ur, lr, ll]
df = pd.DataFrame()
df['lon'] = [100,105,110,115,120,125,130]
df['lat'] = [38,40,42,44,46,48,50]
df['value'] = [1,2,3,4,5,6,7] # not suffice for meshgrid plot
# The colormap to use.
cm = plt.cm.get_cmap('jet')
fig = plt.figure(figsize=[8,6])
ax = plt.axes(projection=ccrs.PlateCarree(), extent=[95, 134, 35, 52])
# plot the red dots using the available data
# comment out if not needed
ax.plot(df['lon'], df['lat'], 'ro')
# plot rectangular patches at the data points
dlon, dlat = 5, 2 #spacings between data points
for lon1, lat1, val1 in zip(df['lon'], df['lat'], df['value']):
pcorners = make_rect(lon1, lat1, dlon, dlat)
poly = mpatches.Polygon(pcorners, ec='gray', fill=True, lw=0.25, \
fc=cm(val1 / max(df['value'])), transform=ccrs.PlateCarree())
ax.add_patch(poly)
ax.gridlines(draw_labels=True)
plt.show()
The output plot:
I can not figure how to make the xticks to display Months.
For a reproducible example, my data is:
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
json = '{"index":{"0":0,"1":1,"2":2,"3":3,"4":4,"5":5,"6":6,"7":7,"8":8,"9":9,"10":10,"11":11,"12":12,"13":13,"14":14,"15":15,"16":16},"Date":{"0":1516147200000,"1":1516752000000,"2":1517788800000,"3":1520208000000,"4":1520985600000,"5":1522281600000,"6":1522886400000,"7":1523404800000,"8":1523491200000,"9":1524096000000,"10":1525305600000,"11":1525737600000,"12":1526428800000,"13":1527811200000,"14":1533686400000,"15":1534377600000,"16":1534809600000},"FB":{"0":0.978943931,"1":1.0282769543,"2":0.999118052,"3":0.994377665,"4":1.0152684601,"5":0.880773866,"6":0.8782934503,"7":0.91676777,"8":0.9032631287,"9":0.9265792518,"10":0.959210704,"11":0.9862198213,"12":1.0098114818,"13":1.0692867773,"14":1.0207253613,"15":0.962958874,"16":0.9514937543},"month":{"0":1,"1":1,"2":2,"3":3,"4":3,"5":3,"6":4,"7":4,"8":4,"9":4,"10":5,"11":5,"12":5,"13":6,"14":8,"15":8,"16":8}}'
toy_data = pd.read_json(json)
fig = plt.figure(figsize=(10,5))
ax = fig.add_subplot(111)
ax.set_title('Share Price Facebook, Google and the SP500')
sns.regplot( x = 'index', y = 'FB', data = toy_data , label = 'FB', fit_reg = True)
plt.show()
I would like to ammend the code so that in the x axis appear the Months of the Range of Observations. That is the tick labels 0 and 1 would be replaced with 'Jan' (appearing once), the tick label 2 would be replaced with 'Feb', the tick labels 3, 4, 5 with 'March' (appearing once), etc.
You can convert your dates to numbers and use those numbers as the x input for the regplot. Then you can format your ticklabels as dates.
import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import seaborn as sns
json = '{"index":{"0":0,"1":1,"2":2,"3":3,"4":4,"5":5,"6":6,"7":7,"8":8,"9":9,"10":10,"11":11,"12":12,"13":13,"14":14,"15":15,"16":16},"Date":{"0":1516147200000,"1":1516752000000,"2":1517788800000,"3":1520208000000,"4":1520985600000,"5":1522281600000,"6":1522886400000,"7":1523404800000,"8":1523491200000,"9":1524096000000,"10":1525305600000,"11":1525737600000,"12":1526428800000,"13":1527811200000,"14":1533686400000,"15":1534377600000,"16":1534809600000},"FB":{"0":0.978943931,"1":1.0282769543,"2":0.999118052,"3":0.994377665,"4":1.0152684601,"5":0.880773866,"6":0.8782934503,"7":0.91676777,"8":0.9032631287,"9":0.9265792518,"10":0.959210704,"11":0.9862198213,"12":1.0098114818,"13":1.0692867773,"14":1.0207253613,"15":0.962958874,"16":0.9514937543}}'
df = pd.read_json(json)
df["Date2"] = mdates.date2num(pd.to_datetime(df["Date"]))
fig = plt.figure(figsize=(10,5))
ax = fig.add_subplot(111)
ax.set_title('Share Price Facebook, Google and the SP500')
sns.regplot( x = 'Date2', y = 'FB', data = df , label = 'FB', fit_reg = True, ax=ax)
loc = mdates.AutoDateLocator()
ax.xaxis.set_major_locator(loc)
ax.xaxis.set_major_formatter(mdates.AutoDateFormatter(loc))
plt.show()
I have an hourly plot generated with matplotlib. I need to plot an event which goes for example, from 09:00 to 10:45. When I try to do it, using axvspan I obtain a bar from 9:00 to 10:00. How could I obtain the longer one?
import matplotlib.pyplot as plt
from matplotlib.gridspec import GridSpec
import datetime as dt
import pandas as pd
now_date = dt.datetime(2018,10,1,9)
d_tw_ini = now_date - dt.timedelta(hours = 1)
d_tw_fin = now_date + dt.timedelta(hours = 3)
dts = pd.date_range(start=d_tw_ini, end=d_tw_fin, freq='1H', name='ini', closed='left')
data=pd.DataFrame({'val':[0.5,0.4,0.7,0.9]})
ev1=[dt.datetime(2018,10,1,9,5),dt.datetime(2018,10,1,10,50)]
data['t']=dts.values
data.set_index('t',inplace=True)
fig = plt.figure()
gs = GridSpec(1, 1)
ax_1 = fig.add_subplot(gs[0, 0])
data.plot(ax=ax_1, y='val')
ax_1.axvspan(ev1[0],ev1[1], alpha=0.3, color= 'red')
Result
Juan, it looks when you used pandas to plot, the hourly indexing seems to cause issues with how axvspan gets plotted.
I replaced
data.plot(ax=ax_1, y='val')
with
ax_1.plot(data.index, data['val'])
which generates the image below, but unfortunately you lose the automated x-axis formatting.
Adding the two lines below will result in the same date formatting as your example.
ax_1.set_xticks([x for x in data.index])
ax_1.set_xticklabels([str(x)[11:16] for x in data.index])
Below is the full code to produce the above plot.
import matplotlib.pyplot as plt
from matplotlib.gridspec import GridSpec
import datetime as dt
import pandas as pd
now_date = dt.datetime(2018,10,1,9)
d_tw_ini = now_date - dt.timedelta(hours = 1)
d_tw_fin = now_date + dt.timedelta(hours = 3)
dts = pd.date_range(start=d_tw_ini, end=d_tw_fin, freq='1h', name='ini',
closed='left')
data=pd.DataFrame({'val':[0.5,0.4,0.7,0.9]})
ev1=[dt.datetime(2018,10,1,9,5,0),dt.datetime(2018,10,1,10,50,0)]
data['t']=dts.values
data.set_index('t',inplace=True)
fig = plt.figure()
gs = GridSpec(1, 1)
ax_1 = fig.add_subplot(gs[0, 0])
# modified section below
ax_1.plot(data.index, data['val'])
ax_1.axvspan(ev1[0],ev1[1], alpha=0.3, color= 'red')
ax_1.set_xticks([x for x in data.index])
ax_1.set_xticklabels([str(x)[11:16] for x in data.index])
plt.show()