vertical_pv.ipynb

vertical_pv.ipynb

First of all, thank you for this Gist.

I have found two issues purely on programming side. I write how usually done below.

1. Slicing on Pandas DataFrame/Series Index: You can use loc.
   (I remember that there used be a warning when loc is not used.)

   e.g. sp.loc[july(4)] instead of sp[july(4):july(4)]
   Also,
   ac_output.loc[july(1):july(5)] instead of ac_output[july(1):july(5)]

2. Avoiding side effect of data modification in Python function: You can deep copy DataFrame inside the function.

   Inside do_the_magic function:
   # This performs deep copy of the DataFrame:
   my_weather_data_copy = my_weather_data.copy()
   # shift the weather times to interval centers to make pandas work more easily
   my_weather_data_copy.index = TIMES
   # Use this copy in the rest of the function
   ...
   # No need to reassign original times to weather_data.index at the end of function

   instead of:
   weather_times = my_weather_data.index
   # shift the weather times to interval centers to make pandas work more easily
   my_weather_data.index = TIMES
   ## The code above causes side effect of modifying weather_data outside the function.
   ## my_weather_data argument passed to the function refers to the same content
   ## as weather_function variable outside the function refers to.
   ## (If you are interested, you can look up how object in Object-Oriented programming
   ## is referenced. Due to the space limitation, I settled with the simple description above.)
   ...
   # ... we stole the
   # references, so now we need to return it to its original state
   weather_data.index = weather_times

   (Note: Under "instead of", comments after # are the original comment and comments after ## are my comments.)

I hope this helps.