Given a tidy Pandas column with 4 or more columns, I want an otherwise very straightforward plot: two of the columns should be the x-y axes of a single figure, and one of the columns should index an Overlay of N Curve objects based on the x-y columns, and N Spread objects, using the final column as error. So if N=4 there should be 4 curves and four spreads. The curves and spreads with same index should be the same color, and the legend should attest to this.
Using table.to(hv.Curve,'col1','col2') I can get a Holomap for the curves, and with some effort I can do the same for the spread. If I then call .overlay() I get a nice figure for the curves including a legend, but when I do the same for the spread the legend vanishes. If I overlay the two, the legend likewise vanishes and the color cycle stops working, making all curves and spreads the same color. If I create a Holomap of curve*spread objects, then the colors match but the legend is still gone.
This seems like a very standard plot, but I can find very little in the Holoviews docs about pairing different Elements or controlling the legend.
This is a bit difficult to answer without any concrete code, for example I can't reproduce some of the issues you are describing. However the first issue is simply that show_legend is not enabled by default for the Spread elemen. In the case of plotting a Curve and Spread using .to and .overlay, here is what I can confirm works:
%%opts Spread [show_legend=True width=600] Overlay [legend_position='right']
df = pd.DataFrame({
'index': np.arange(100), 'y': np.random.randn(100).cumsum(),
'err': np.random.rand(100)+0.1, 'z': np.repeat(np.arange(10), 10)
})
ds = hv.Dataset(df)
ds.to(hv.Curve, 'index', 'y', 'z').overlay() * ds.to(hv.Spread, 'index', ['y', 'err']).overlay()
If I create a Holomap of curve*spread objects, then the colors match but the legend is still gone.
This is indeed a current limitation since we recommended against nesting objects in this way in the past, however I have just opened this PR which will allow this approach as well.
Related
I'm trying to use Seaborn to plot two curves that shouldn't overlap perfectly. The blue curve in this example should be defined from 2 to 8, and the red line should be defined from 1 to 7, so they should appear slightly staggered. However, sns.pointplot draws both curves as if they overlapped perfectly. Here is the most minimal code I used to generate my plot:
g = sns.FacetGrid(myDF, col = 'condition', hue = 'manipulation',
col_order = ['condition1', 'condition2', 'condition3'])
g.map(sns.pointplot, 'ordinalIV', 'continuousDV', **{'x_estimator': np.mean})
Is there any way to shift these curves, so that they don't overlap? This issue seems similar to my problem, but I want to keep the bootstrapped confidence intervals in my plot, so I'm not sure how to re-create that. If anyone knows of a workaround to (1) compute bootstrapped confidence intervals outside of seaborn and (2) add them to a matplotlib plot, that would work too!
I am really pretty new to matplotlib, though I know that it can be very powerful.
I've been reading number of tutorials and examples and it's a real hassle to understand how does matplotlib's Figure and Axes work. I am illustrating, what I am trying to understand, with the attached figure.
I know how to create a figure instance of certain size in inches. However, what bothers me is how can I create subplots and then axes, within each subplot, with relative coordinates (bottom=0,left=0,top=1,right=1) as illustrated.
So, for example I want to create a "parent" plot area (say (6in,10in)). Then, I want to create two subplot areas, each with size (3in,3in), with 1in space from the top, 2in space between the two vertical subplot areas and 1in from bottom. Then, 1in space on the left and 2in space on the write. In the same time, I would like to be able to get the coordinates of the subplot areas with respect to the main plot area.
Then, inside the first subplot area, I'd like to create 2 axis instances, with Axis 1, having coordinates with respect to Subplot Area1 (0.1,0.7,0.7,0.2) and Axes 2 (0.1,0.2,0.7,0.5). And then of course I'd like to be able to plot on these axes e.g., ax1.plot()....
If you could provide a sample code to achieve that, then I can study it.
Your help will be very much appreciated!
a subplot and an Axes object are really the same thing. There is not really a "subplot" as you describe it in matplotlib. You can just create your three Axes objects using gridspec without the need to put them in your "subplots".
There are a few different ways to create Axes instances within your figure.
fig.add_axes will create an Axes instance at the position given to it (you give it [left,bottom,width,height] in figure coordinates (i.e. 0,0 is bottom left, 1,1 is top right).
fig.add_subplot will also create an Axes instance. In this case, rather than giving it a rectangle to be created in, you give it the number of rows and columns of subplots you would like, and then the plot_number, where plot_number starts at 1, increments across rows first and has a maximum of nrows * ncols.
For example, to create the top-left Axes in a grid of 2 row and 2 columns, you could do the following:
fig.add_subplot(2,2,1)
or the shorthand
fig.add_subplot(221)
There are some more customisable ways to create Axes as well, for example gridspec and subplot2grid which allow for easy creation of many subplots of different shapes and sizes.
I'm trying to get and set the position of a draggable legend in matplotlib. My application consists of an interactive GUI, which has a redraw/plot function that should perform the follow steps:
save the position of the current legend.
clear the current axes and perform various plotting operations, which may or may add labels to their plots.
build a new draggable legend (ax.legend().draggable()) and restore the old position of the legend.
In between these steps the user is free to drag the legend around, and the goal is to persist the legend position when the plots are redrawn.
My first approach was to use oldpos = legend.get_bbox_to_anchor() and legend.set_bbox_to_anchor(oldpos) in steps 1 and 3. However this causes to move the legend completely off the visible area.
Note that I have to use ax.legend() and cannot use fig.legend(lines, labels), since step 2 is completely decoupled, i.e., I don't know anything about lines and labels in step 3. According to answers to the question How to position and align a matplotlib figure legend? there seems to be a difference between these two possibilities regarding axes or figure coordinates. Obviously my problem calls for figure coordinates, but I haven't fully understood how to convert the bbox to a "bbox in figure coordinates".
The even more severe problem I just realized is that apparently legend.get_bbox_to_anchor() always seems to return the same values irrespective of the drag position. So maybe the anchor can only be (ab-)used to manipulate the position of static legends? Is there another/proper way to save and restore the position of a draggable legend?
By looking at the implementation of Legend I found out that there is an undocumented property _loc, which exactly does what I want. My solution now looks astonishingly simple:
oldLegPos = ax.get_legend()._loc
# perform all plotting operations...
legend = ax.legend().draggable()
legend._loc = oldLegPos
It looks like _loc automatically stores figure coordinates, since I do not have to convert the coordinates in any way (eg. when the plotting operations completely change the axes ranges/coordinates).
I have images that have had features extracted with a contouring algorithm (I'm doing astrophysical source extraction). This approach yields a "feature map" that has each pixel "labeled" with an integer (usually ~1000 unique features per map).
I would like to show each individual feature as its own contour.
One way I could accomplish this is:
for ii in range(labelmask.max()):
contour(labelmask,levels=[ii-0.5])
However, this is very slow, particularly for large images. Is there a better (faster) way?
P.S.
A little testing showed that skimage's find-contours is no faster.
As per #tcaswell's comment, I need to explain why contour(labels, levels=np.unique(levels)+0.5)) or something similar doesn't work:
1. Matplotlib spaces each subsequent contour "inward" by a linewidth to avoid overlapping contour lines. This is not the behavior desired for a labelmask.
2. The lowest-level contours encompass the highest-level contours
3. As a result of the above, the highest-level contours will be surrounded by a miniature version of whatever colormap you're using and will have extra-thick contours compared to the lowest-level contours.
Sorry for answering my own... impatience (and good luck) got the better of me.
The key is to use matplotlib's low-level C routines:
I = imshow(data)
E = I.get_extent()
x,y = np.meshgrid(np.linspace(E[0],E[1],labels.shape[1]), np.linspace(E[2],E[3],labels.shape[0]))
for ii in np.unique(labels):
if ii == 0: continue
tracer = matplotlib._cntr.Cntr(x,y,labels*(labels==ii))
T = tracer.trace(0.5)
contour_xcoords,contour_ycoords = T[0].T
# to plot them:
plot(contour_xcoords, contour_ycoords)
Note that labels*(labels==ii) will put each label's contour at a slightly different location; change it to just labels==ii if you want overlapping contours between adjacent labels.
I'm trying to display 2D data with axis labels using both contour and pcolormesh. As has been noted on the matplotlib user list, these functions obey different conventions: pcolormesh expects the x and y values to specify the corners of the individual pixels, while contour expects the centers of the pixels.
What is the best way to make these behave consistently?
One option I've considered is to make a "centers-to-edges" function, assuming evenly spaced data:
def centers_to_edges(arr):
dx = arr[1]-arr[0]
newarr = np.linspace(arr.min()-dx/2,arr.max()+dx/2,arr.size+1)
return newarr
Another option is to use imshow with the extent keyword set.
The first approach doesn't play nicely with 2D axes (e.g., as created by meshgrid or indices) and the second discards the axis numbers entirely
Your data is a regular mesh? If it doesn't, you can use griddata() to obtain it. I think that if your data is too big, a sub-sampling or regularization always is possible. If the data is too big, maybe your output image always will be small compared with it and you can exploit this.
If you use imshow() with "extent" and "interpolation='nearest'", you will see that the data is cell-centered, and extent provided the lower edges of cells (corners). On the other hand, contour assumes that the data is cell-centered, and X,Y must be the center of cells. So, you need to be care about the input domain for contour. The trivial example is:
x = np.arange(-10,10,1)
X,Y = np.meshgrid(x,x)
P = X**2+Y**2
imshow(P,extent=[-10,10,-10,10],interpolation='nearest',origin='lower')
contour(X+0.5,Y+0.5,P,20,colors='k')
My tests told me that pcolormesh() is a very slow routine, and I always try to avoid it. griddata and imshow() always is a good choose for me.