How to keep/assign node names when contracting vertices in igraph? - python

I am trying to contract some vertices in igraph (using the python api) while keeping the names of the vertices. It isn't clear to me how to keep the name attribute of the graph. The nodes of the graph are people and I'm trying to collapse people with corrupted names.
I looked at the R documentation and I still don't see how to do it.
For example, if I do either of the following I get an error.
smallgraph.contract_vertices([0,1,2,3,4,2,6],vertex.attr.comb=[name='first'])
smallgraph.contract_vertices([0,1,2,3,4,2,6],vertex.attr.comb=['first'])

In Python, the keyword argument you need is called combine_attrs and not vertex.attr.comb. See help(Graph.contract_vertices) from the Python command line after having imported igraph. Also, the keyword argument accepts either a single specifier (such as first) or a dictionary. Your first example is invalid because it is simply not valid Python syntax. The second example won't work because you pass a list with a single item instead of just the single item.
So, the correct variants would be:
smallgraph.contract_vertices([0,1,2,3,4,2,6], combine_attrs=dict(name="first"))
smallgraph.contract_vertices([0,1,2,3,4,2,6], combine_attrs="first")

Nevermind. You can just enter a dictionary without using the wording
vertex.attr.comb

Related

Instancing maya objects with a sequential suffix, object name string not seen by cmds.instance

I have a question about string usage in lists in python for Maya. I am writing a script meant to take a selected object, then instance it 100 times with random translate, scale, and orient attributes. The script itself works and does what it's meant to, however I'm not being able to decipher how to instance the objects with the original object name, and then add a suffix that ends with "_instance#", where # assigns 1, 2, 3, etc. in order to the copies of the original mesh. This is where I'm at so far:
#Capture selected objects, sort into list
thing = MC.ls(sl=True)
print thing
#Create instances of objects
instanceObj = MC.instance(thing, name='thing' + '_instance#')
This returns a result that looks like "thing_instance1, thing_instance2".
Following this, I figured the single quote around the string for the object was causing it to just name it "thing", so I attempted to write it as follows
MC.instance(thing, name=thing + '_instance1'
I guess because instance uses a list, it's not accepting the second usage of the string as valid and returns a concatenate error. I've tried rewriting this a few times and the closest I get is with
instanceObj = MC.instance(thing)
which results in a list of (pCube1,2,3,4), but is lacking the suffix.
I'm not sure where to go from here to end up with a result where the instanced objects are named with the convention "pCube1_instance1, pCube1_instance2" etc.
Any assistance would be appreciated.
It is not clear if you want to use only one source object or more. In any case the
MC.ls(sl=True)
returns a list of strings. And concatenating a list and a string does not work. So use thing[0] or simply
MC.ls(sl=True)[0]
If you get errormessages, please always include the message in your question, it helps a lot to see what error appears.

(graph-tool) how to get the value from property map as an array

First, I'm newbie on python.
This is my code to get centrality betweenness using graph-tool module on python3
vertex_betweenness = betweenness(g)
then after I got a property map which contains the value I want.
I have tried this
for v in g.vertices():
print(vertex_betweenness[v])
I have tried to output it in some ways, print to the screen or create csv file but I cannot do it. I think I don't know about how to iterate through that property map.
I already googled for a few days but no pages can help me solve this question.
The function betweenness() returns two property maps, for both vertex and edge property maps (see the documentation at https://graph-tool.skewed.de/static/doc/centrality.html#graph_tool.centrality.betweenness). Your snipped will work if you modify it to
vertex_betweenness, edge_betweenness = betweenness(g)
In the docs, you can get the scalar values from a property map using the .get_array() method or .a. Seems to be the easiest way for me!

How to parse for length of functions in a python module?

Python modules can be parsed with ast.parse.
With the module object returned, nodes can be iterated and attributes like lineno and name can be accessed.
Currently, the way I use to determine the length (number of lines) of a function is to find the node X that corresponds to the function, and get the node Y right after it, then Y.lineno - X.lineno is the length (possibly with blank lines included).
But this method is not fast enough to me. Is there any tools that I can use to do this kind of static analysis?

error with gdalbuildvrt, in Python

I am new to python/GDAL and am running into perhaps a trivial issue. This may stem from the fact that I don't really understand how to use GDAL properly in python, or something careless, but even though I think I am following the help doc, I keep getting a syntax error when trying to use "gdalbuildvrt".
What I want to do is take several (amount varies for each set, call it N) geotagged 1-band binary rasters [all values are either 0 or 1] of different sizes (each raster in the set overlaps for the most part though), and "stack" them on top of each other so that they are aligned properly according to their coordinate information. I want this "stack" simply so I can sum the values and produce a 'total' tiff that has an extent to match the exclusive extent (meaning not just the overlap region) of all the original rasters. The resulting tiff would have values ranging from 0 to N, to represent the total number of "hits" the pixel in that location received over the course of the N rasters.
I was led to gdalbuildvrt [http://www.gdal.org/gdalbuildvrt.html] and after reading about it, it seemed that by using the keyword -separate, I would be able to achieve what I need. However, each time I try to run my program, I get a syntax error. The following shows two of the several different ways I tried calling gdalbuildvrt:
gdalbuildvrt -separate -input_file_list stack.vrt inputlist.txt
gdalbuildvrt -separate stack.vrt inclassfiles
Where inputlist.txt is a text file with a path to the tif on every line, just like the help doc specifies. And inclassfiles is a python list of the pathnames. Every single time, no matter which way I call it, I get a syntax error on the first word after the keywords (i.e. 'inputlist' in inputlist.txt, or 'stack' in stack.vrt).
Could someone please shed some light on what I might be doing wrong? Alternatively, does anyone know how else I could use python to get what I need?
Thanks so much.
gdalbuildvrt is a GDAL command line utility. From your example its a bit unclear how you actually run it, but when running from within Python you should execute it as a subprocess.
And in your first line you have the .vrt and the .txt in the wrong order. The textfile containing the files should follow directly after the -input_file_list.
From within Python you can call gdalbuildvrt like:
import os
os.system('gdalbuildvrt -separate -input_file_list inputlist.txt stack.vrt')
Note that the command is provided as a string. Using a Python list with the files can be done with something like:
os.system('gdalbuildvrt -separate stack.vrt %s') % ' '.join(data)
The ' '.join(data) part converts the list to a string with a space between the items.
Depending on how your GDAL is build, its sometimes possible to use wildcards as well:
os.system('gdalbuildvrt -separate stack.vrt *.tif')

Reducing menu.add_command() clutter/repeat lines

I would like to do the following (just an example, the real code has more menu's and more add_command's):
editmenu.add_command(label="Cut",state="disabled")
editmenu.add_command(label="Copy",state="disabled")
editmenu.add_command(label="Paste",state="disabled")
editmenu.add_command(label="Delete",state="disabled")
But on fewer lines, In fact, just one line if possible. I have menus that are taking up a considerable amount of space in my program and would like to reduce the clutter. Plus the programmer in me sees a bunch of similar lines and feels there must be a way to reduce them.
I tried the following code to no avail; I obviously got a nameerror because label and state aren't defined...
for labeldic in [{label:"Cut"},{label:"Copy"},{label:"Paste"},{label:"Delete"}]: editmenu.add_command(labeldic+{state:"disabled"})
Thanks in advance for any suggestions!
Here's a translation of what you wanted to do:
for labeldic in [{"label":"Cut"},{"label":"Copy"},{"label":"Paste"},{"label":"Delete"}]:
labeldic.update({"state": "disabled"})
editmenu.add_command(**labeldic)
There were three problems I fixed.
The first is that dictionary keys need to be quoted if they are strings. If you want a dict mapping the string 'label' to the the string 'cut', you can do it using the dict literal {'label': 'cut'}, or else possibly with the dict() constructor which expands keyword arguments that way: dict(label='cut'). As you discovered, {label: 'cut'} wouldn't work, because it tries to use a variable's value for the key, but there is no such variable.
The second is that you can't merge dictionaries using the + operator. It doesn't work, unfortunately. There is, however, an update method that mutates the dict it's called on. Since it doesn't return a merged dict, it can't be used inline the way you used +.
The third problem is that passing a dict is not the same as passing in keyword arguments. foo(bar='baz') is not the same as foo({'bar':'baz'}), but it is the same as foo(**{'bar':'baz'}). The ** syntax in function calling "unpacks" a dictionary into keyword arguments.
Regardless it's sort of weird style. Here's what I would do instead:
for label in ['Cut', 'Copy', 'Paste', 'Delete']:
editmenu.add_command(label=label, state='disabled')

Categories

Resources