Pandas FutureWarning in westeros_historical_new_capacity.ipynb

[Tyler-lc]

What is this about?

Deprecated method in Pandas
In the tutorial hystorical_new_capacity we use the command df["value"] = float(df[df["year"] == 700]["value"]) modify the demand.
The problem is that this will throw a warning because it will be deprecated in future version of Pandas:

:1: FutureWarning: Calling float on a single element Series is deprecated and will raise a TypeError in the future. Use float(ser.iloc[0]) instead

I suggest to replace the above line with either a lengthy:

filter_df = df["year"] == 700
value = df[filter_df]["value"].values[0]
df["value"] = value

or a more compact:
df["value"] = df.loc[df["year"] == 700, "value"].iloc[0]

since the values in the column are already numpy.float64 it should not be a problem. If for some reason we really need to make sure that it is a float we could always use
df["value"] = df["year"].apply(float)

Grid Efficiency value
In this exercise we need to retrieve the grid_efficiency value so that we can calculate how much coal_ppl capacity we need. The following is used:

# We know that there are losses that occur when electricity is transmitted via the
# `grid`.
grid_eff = float(
    scenario.par(
        "output",
        filters={
            "year_vtg": scenario.firstmodelyear,
            "year_act": scenario.firstmodelyear,
            "technology": "grid",
        },
    )["value"]
)
print(f"Grid efficiency is {grid_eff}.")
demand_of_firstmodelyear /= grid_eff

This will give 1 as a result because the output of the grid is 1.
I think we are looking for the input which is 1.111111_ in which case the correct code would be:

grid_eff = float(
    scenario.par(
        "input",
        filters={
            "year_vtg": scenario.firstmodelyear,
            "year_act": scenario.firstmodelyear,
            "technology": "grid",
        },
    )["value"]
)

[glatterf42]

I'm not sure I agree on the grid_efficiency part of this issue. In my tutorial notebook, when I execute this cell,

grid_eff = float(
    scenario.par(
        "output",
        filters={
            "year_vtg": scenario.firstmodelyear,
            "year_act": scenario.firstmodelyear,
            "technology": "grid",
        },
    )["value"].iloc[0]
)
print(f"Grid efficiency is {grid_eff}.")
demand_of_firstmodelyear /= grid_eff

I see this output:

Grid efficiency is 0.9.

Which is exactly what we said in the westeros_baseline tutorial. We are using this value to increase the demand_of_firstmodelyear to reflect that we need higher input by 1/grid_efficiency so that the demanded output is achieved.

[Tyler-lc]

When I run:

grid_eff = float(
    scenario.par(
        "output",
        filters={
            "year_vtg": scenario.firstmodelyear,
            "year_act": scenario.firstmodelyear,
            "technology": "grid",
        },
    )["value"].iloc[0]
)
print(f"Grid efficiency is {grid_eff}.")
demand_of_firstmodelyear /= grid_eff

I get that the grid_efficiency is 1.0.

When looking at the westeros_baseline from the official repo we have that:

grid_efficiency = 0.9
year_df = scenario.vintage_and_active_years((country, "grid"), in_horizon=False)
vintage_years, act_years = year_df["year_vtg"], year_df["year_act"]

base.update(
    year_vtg=vintage_years,
    year_act=act_years,
    technology="grid",
    commodity="electricity",
    level="final",
    value=1.0,
)
grid_out = make_df("output", **base, node_dest=country, time_dest="year")
scenario.add_par("output", grid_out)

base.update(commodity="electricity", level="secondary", value=1.0 / grid_efficiency)
grid_in = make_df("input", **base, node_origin=country, time_origin="year")
scenario.add_par("input", grid_in)

We are setting the value = 1.0/grid_efficiency on the input parameter. As you can see here the output is set to 1.0.
I don't know why we get different numbers though. That is interesting 😄