I defined a hello world function in a file called 'functions.ipynb'. Now, I would like to import functions in another file by using "import functions". I am sure that they are in the same folder. However, it still shows that "ImportError: No module named functions". By the way, I am using jupyter notebook. Thanks a lot!
You'll want to use the ipynb package/module importer. You'll need to install it: pip install ipynb.
Create a Notebook named my_functions.ipynb. Add a simple function to it.
def factorial(n):
if n == 0:
return 1
else:
return n * factorial(n-1)
Then, create a second IPython Notebook and import this function with:
from ipynb.fs.full.my_functions import factorial
Then you can use it as if it was in the same IPython Notebook:
testing = factorial(5)
See the documentation for more details.
For my use case the ipnyb import didn't work for some reason. I had to use Jupyter Notebook magic cell to import my function.:
%run MyOtherNotebook.ipynb #this is were my function was stored
function(df) #then simply run the function
#David Rinck's answer solved the problem, but I'd like to recommend you add the boilerplate __name__ == "__main__" to guard the scripts you don't want to accidentally invoke. It works the same way as in a usual Python file.
If a .ipynb file a.ipynb is imported by another one b.ipynb
from ipynb.fs.full.a import factorial
the __name__ in a.ipynb would be ipynb.fs.full.a rather than "__main__".
You can save functions.ipynb as functions.py and can import the file as import functions. Now you can use any function defined in the functions file as functions.function_name
For eg, if add is a function,
functions.add(5,3)
after importing will work.
Related
I'm trying to use multiple files in my programing and I ran into a problem.
I have two files: main.py, nu.py
The main file is:
import numpy
import nu
def numarray():
numpy.array(some code goes here)
nu.createarray()
The nu file is:
def createarray():
numpy.array(some code goes here)
When I run main I get an error:
File "D:\python\nu.py", line 2, in createarray
numpy.array(some code goes here)
NameError: name 'numpy' is not defined
numpy is just an exaple, I'm using about six imports.
As far as I see it, I have to import all moduls on all files, but it creating a problem where certain modules can't be loaded twice, it just hang.
What I'm doing wrong and how can I properly import functions from another file while using imported modules from the main file?
I hope i explain it well.
thanks for helping!
I have years in python and importing from other files is still a headache..
The problmen here is that you are not importing numpy in "nu.py".
But as you say sometimes it's a little annoying have to import al libraries in all files.
The last thing is, How do you get the error a module cannot be imported twice? can you give me an example?
In each separate python script if you are using a module within you need to import it to access. So you will need to 'import numpy' in your nu.py script like below
If possible try keeping the use of a module within a script so you dont have import the same multiple times, although this wont always be appropriate
I am using Python 2.7 on Windows 7 Professional.
I am trying to call a function saved in another file to run in this file's code.
Function called dosomething is found in anotherfile.py
anotherfile.py is in the same directory as current code.
My call in this file is simple:
import anotherfile
print anotherfile.dosomething
I am getting an error: No module named anotherfile
The problem is the same as I found in this post
I don't understand the solution but I'd like any insight?
Thank you.
EDIT: The other question/answers discuss resetting CLASSPATH and setting PYTHONPATH. I explored this but was not sure how to do this. Perhaps relevant?
Let us have two files in the same directory. Files are called main.py and another.py.
First write a method in another.py:
def do_something():
return "This is the do something method"
Then call the method from main.py. Here is the main.py:
import another
print another.do_something()
Run main.py and you will get output like this:
This is the do something method
N.B.: The above code is being executed using Python 2.7 in Windows 10.
Specify the module then the file then the import like so:
from this_module.anotherfile import dosomething
or if you want all functions from "anotherfile.py"
from this_module.anotherfile import *
and then you can call the "dosomething" command without the "anotherfile" prefix.
I ran into same problem. After ample of trials, I ended up solving it with the below mentioned solution:
Make sure your current file and anotherfile.py lies in same location of system path.
Say your another.py and current file lies at location : "C:/Users/ABC"
In case, one is not aware of system path. Use below code in current file:
import sys
print(sys.path)
import sys
sys.path.append('/C:/Users/ABC/')
Then you do below code in same current code:
from another import dosomething
I found the issue. Python was looking in another directory for the files. I explicitly set my working directory as the path to where thisfile.py and anotherfile.py reside and it works. Thank you for all the quick replies.
So, I've two python files:
the 1st "m12345.py"
def my():
return 'hello world'
the 2nd "1234.py":
from m12345 import *
a = m12345.my()
print(a)
On ipython I try to exec such cmds:
exec(open("f:\\temp\\m12345.py").read())
exec(open("f:\\temp\\1234.py").read())
the error for the 2nd command is:
ImportError: No module named 'm12345'
Please, help how to add the 1st file as a module for the 2nd?
First off, if you use the universal import (from m12345 import *) then you just call the my() function and not the m12345.my() or else you will get a
NameError: name 'm12345' is not defined
Secondly, you should add the following snippet in every script in which you want to have the ability of directly running it or not (when importing it).
if "__name__" = "__main__":
pass
PS. Add this to the 1st script ("m12345.py").
PS2. Avoid using the universal import method since it has the ability to mess the namespace of your script. (For that reason, it isn't considered best practice).
edit: Is the m12345.py located in the python folder (where it was installed in your hard drive)? If not, then you should add the directory it is located in the sys.path with:
import sys
sys.path.append(directory)
where directory is the string of the location where your m12345.py is located. Note that if you use Windows you should use / and not \.
However it would be much easier to just relocate the script (if it's possible).
You have to create a new module (for example m12345) by calling m12345 = imp.new_module('m12345') and then exec the python script in that module by calling exec(open('path/m12345.py').read(), m12345.__dict__). See the example below:
import imp
pyfile = open('path/m12345.py').read()
m12345 = imp.new_module('m12345')
exec(pyfile, m12345.__dict__)
If you want the module to be in system path, you can add
sys.modules['m12345'] = m12345
After this you can do
import m12345
or
from m12345 import *
I am working on project euler and wanted to time all of my code. What I have is directory of files in the form 'problemxxx.py' where xxx is the problem number. Each of these files has a main() function that returns the answer. So I have created a file called run.py, located in the same directory as the problem files. I am able to get the name of the file through command prompt. But when I try to import the problem file, I continue to get ImportError: No module named problem. Below is the code for run.py so far, along with the command prompt used.
# run.py
import sys
problem = sys.argv[1]
import problem # I have also tired 'from problem import main' w/ same result
# will add timeit functions later, but trying to get this to run first
problem.main()
The command prompts that I have tried are the following: (both of which give the ImportError stated above)
python run.py problem001
python run.py problem001.py
How can I import the function main() from the file problem001.py? Does importing not work with the file name stored as a variable? Is there a better solution than trying to get the file name through command prompt? Let me know if I need to add more information, and thank you for any help!
You can do this by using the __import__() function.
# run.py
import sys
problem = __import__(sys.argv[1], fromlist=["main"]) # I have also tired 'from problem import main' w/ same result
problem.main()
Then if you have problem001.py like this:
def main():
print "In sub_main"
Calling python run.py problem001 prints:
In sub_main
A cleaner way to do this (instead of the __import__ way) is to use the importlib module. Your run.py needs to changes:
import importlib
problem = importlib.import_module(sys.argv[1])
Alternatives are mentioned in this question.
For sure! You can use __ import_ built-in function like __import__(problem). However this is not recommended to use, because it is not nice in terms of coding-style. I think if you are using this for testing purposes then you should use unittest module, either way try to avoid these constructions.
Regards
You can use exec() trick:
import sys
problem = sys.argv[1]
exec('import %s' % problem)
exec('%s.main()' % problem)
I've been trying to evaluate a simple "integrate(x,x)" statement from within Python, by following the Sage instructions for importing Sage into Python. Here's my entire script:
#!/usr/bin/env sage -python
from sage.all import *
def main():
integrate(x,x)
pass
main()
When I try to run it from the command line, I get this error thrown:
NameError: global name 'x' is not defined
I've tried adding var(x) into the script, global x, tried replacing integrate(x,x) with sage.integrate(x,x), but I can't seem to get it to work, I always get an error thrown.
The command I'm using is ./sage -python /Applications/path_to/script.py
I can't seem to understand what I'm doing wrong here.
Edit: I have a feeling it has something to do with the way I've "imported" sage. I have my a folder, let's call it folder 1, and inside of folder 1 is the "sage" folder and the "script.py"
I am thinking this because typing "sage." doesn't bring up any autocomplete options.
The name x is not imported by import sage.all. To define a variable x, you need to issue a var statement, like thus
var('x')
integrate(x,x)
or, better,
x = SR.var('x')
integrate(x,x)
the second example does not automagically inject the name x in the global scope, so that you have to explicitly assign it to a variable.
Here's what Sage does (see the file src/sage/all_cmdline.py):
from sage.all import *
from sage.calculus.predefined import x
If you put these lines in your Python file, then integrate(x,x) will work. (In fact, sage.calculus.predefined just defines x using the var function from sage.symbolic.ring; this just calls SR.var, as suggested in the other answer. But if you want to really imitate Sage's initialization process, these two lines are what you need.)