# Frequently Answered Questions / Troubleshooting¶

## LaTeX & Waf¶

### 'error when calling biber, check xxx.blg for errors'¶

This is a well-known bug in biber that occurs occasionally. Nicely explained here.

Short excerpt from LaTeX Stack Exchange for the fix:

You need to delete the relevant cache folders and compile your document again. You can find the location of the cache folder by looking at the .blg file, or by using the command:

biber --cache


On Linux and Mac, this can be combined to delete the offending folder in one command:

rm -rf biber --cache


In my experience, it helps to run waf on only one core for the first time you compile multiple LaTeX documents (once Biber’s cache is built correctly, you can do this in parallel again):

python waf.py -j1


### Biber on 64-bit MikTeX¶

There have been mulitple issues of with biber on Windows, sometimes leading to strange error messages from Python’s subprocess module (e.g., “file not found” errors). Apparently, current 64-bit MikTeX distributions do not contain Biber. Two possible fixes:

• Recommended: In the main wscript file, replace the line:

ctx.load('biber')


by:

ctx.load('tex')


In src/paper/research_paper.tex and src/paper/research_pres_30min.tex, replace:

backend=biber


by:

backend=bibtex

• For the adventurous: Download a 64-bit version of biber here: http://biblatex-biber.sourceforge.net/. Put the file biber.exe into the correct folder, typically that will be “C:Program FilesMiKTeX 2.9miktexbinx64”, “C:Program FilesMiKTeX 2.9miktexbin”, or the like. Hat tip to Andrey Alexandrov, a student in my 2014 class at Bonn.

## Using Spyder with Waf¶

Spyder is a useful IDE for developing scientific Python code – it has been specifically developed for this purpose and has first-class support for data structures like NumPy arrays and pandas dataframes.

In the context of these project templates, there are just two issues to consider.

### Debugging wscript files¶

Sometimes it is helpful to debug your build code directly. As the wscript files are just pure Python code Spyder can handle them in principle. The tricky bit is to make Spyder recognise the them as Python scripts – usually it just uses the extension .py to infer that fact. As you cannot simply add this extension to wscript files, you must tell Spyder inside the wscript file using a so-called “shebang”. Simply add the following line as the first thing to all your wscripts:

#! python


You can then set breakpoints inside your wscript files and debug them by running waf.py from inside spyder (just make sure you ran python waf.py configure beforehand).

### Setting the PYTHONPATH¶

The machinery of the imports in Python scripts requires the PYTHONPATH environmental variable to include the project root; you will need to add the project root directory to the PYTHONPATH when debugging files in Spyder as well.

In order to do so, first create a Spyder project in the directory where you want your research to be (click “Yes” on the question “The following directory is not empty: ... Do you want to continue?”). Then right-click on the project’s root folder and select “Add to PYTHONPATH”.

Any ImportErrors are likely due to this not being done correctly. Note that you must set the run configuration (F6 or select “Run” from the menu bar and then “Configure”) to “Execute in a new dedicated Python console”.

## Stata packages¶

Note that when you include (or input) the file project_paths.do in your Stata script, the system directories get changed. This means that Stata will not find any packages you installed system-wide anymore. This is desired behaviour to ensure that you (and your coauthors) run the same versions of different packages that you installed via ssc or the like. The project template comes with a few of them, see src/library/stata/ado_ext in the stata branch.

1. Open a Stata command line session and change to the project root directory
2. Type include bld/project_paths
3. Type sysdir and make sure that the PLUS and PERSONAL directories point to subdirectories of the project.
4. Install your package via ssc, say ssc install tabout

## Stata failure: FileNotFoundError¶

The following failure:

[21/39] Running  [Stata] -e -q do src/data_management/add_variables.do add_variables
Waf: Leaving directory /Users/xxx/econ/econ-project templates/bld'
Build failed
Traceback (most recent call last):
File "/Users/xxx/econ/econ-project templates/.mywaflib/waflib/Task.py", line 212, in process
ret = self.run()
File "/Users/xxx/econ/econ-project templates/.mywaflib/waflib/extras/run_do_script.py", line 140, in run
ret, log_tail = self.check_erase_log_file()
File "/Users/xxx/econ/econ-project templates/.mywaflib/waflib/extras/run_do_script.py", line 166, in check_erase_log_file
with open(**kwargs) as log:
FileNotFoundError: [Errno 2] No such file or directory: '/Users/xxx/econ/econ-project templates/bld/add_variables.log'


has a simple solution: Get rid of all spaces in the path to the project. (i.e., econ-project-templates instead of econ-project templates in this case). To do so, do not rename your user directory, that will cause havoc. Rather move the project folder to a different location.

I have not been able to get Stata working with spaces in the path in batch mode, so this has nothing to do with Python/Waf. If anybody finds a solution, please let me know.

## Stata failure: missing file¶

If you see this error:

[21/39] Running  [Stata] -e -q do src/data_management/add_variables.do add_variables
Waf: Leaving directory /Users/xxx/econ/econ-project/templates/bld'
Build failed

run python waf.py configure again and check that you have a license for the Stata version that is found (the Stata tool just checks availability top-down, i.e., MP-SE-IC, in case an MP-Version is found and you just have a license for SE, Stata will silently refuse to start up).