I want to get a list of the data contained in a histogram bin. I am using numpy, and Matplotlib. I know how to traverse the data and check the bin edges. However, I want to do this for a 2D histogram and the code to do this is rather ugly. Does numpy have any constructs to make this easier?
For the 1D case, I can use searchsorted(). But the logic is not that much better, and I don’t really want to do a binary search on each data point when I don’t have to.
Most of the nasty logic is due to the bin boundary regions. All regions have boundaries like this: [left edge, right edge). Except the last bin, which has a region like this: [left edge, right edge].
Here is some sample code for the 1D case:
import numpy as np
data = [0, 0.5, 1.5, 1.5, 1.5, 2.5, 2.5, 2.5, 3]
hist, edges = np.histogram(data, bins=3)
print 'data =', data
print 'histogram =', hist
print 'edges =', edges
getbin = 2 #0, 1, or 2
print '---'
print 'alg 1:'
#for i in range(len(data)):
for d in data:
if d >= edges[getbin]:
if (getbin == len(edges)-2) or d < edges[getbin+1]:
print 'found:', d
#end if
#end if
#end for
print '---'
print 'alg 2:'
for d in data:
val = np.searchsorted(edges, d, side='right')-1
if val == getbin or val == len(edges)-1:
print 'found:', d
#end if
#end for
Here is some sample code for the 2D case:
import numpy as np
xdata = [0, 1.5, 1.5, 2.5, 2.5, 2.5, \
0.5, 0.5, 0.5, 0.5, 1.5, 1.5, 1.5, 1.5, 1.5, 2.5, 2.5, 2.5, 2.5, 2.5, 2.5, \
0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 1.5, 1.5, 1.5, 1.5, 1.5, 1.5, 1.5, 1.5, 2.5, 2.5, 2.5, 2.5, 2.5, 2.5, 2.5, 2.5, 3]
ydata = [0, 5,5, 5, 5, 5, \
15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, \
25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 30]
xbins = 3
ybins = 3
hist2d, xedges, yedges = np.histogram2d(xdata, ydata, bins=(xbins, ybins))
print 'data2d =', zip(xdata, ydata)
print 'hist2d ='
print hist2d
print 'xedges =', xedges
print 'yedges =', yedges
getbin2d = 5 #0 through 8
print 'find data in bin #', getbin2d
xedge_i = getbin2d % xbins
yedge_i = int(getbin2d / xbins) #IMPORTANT: this is xbins
for x, y in zip(xdata, ydata):
# x and y left edges
if x >= xedges[xedge_i] and y >= yedges[yedge_i]:
#x right edge
if xedge_i == xbins-1 or x < xedges[xedge_i + 1]:
#y right edge
if yedge_i == ybins-1 or y < yedges[yedge_i + 1]:
print 'found:', x, y
#end if
#end if
#end if
#end for
Is there a cleaner / more efficient way to do this? It seems like numpy would have something for this.
Click Data > Data Analysis > Histogram > OK. Under Input, select the input range (your data), then select the bin range.
The histogram is a popular graphing tool. It is used to summarize discrete or continuous data that are measured on an interval scale.
digitize
, from core NumPy, will give you the index of the bin to which each value in your histogram belongs:
import numpy as NP
A = NP.random.randint(0, 10, 100)
bins = NP.array([0., 20., 40., 60., 80., 100.])
# d is an index array holding the bin id for each point in A
d = NP.digitize(A, bins)
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With