Pandas and Ibis API

Functions in the era_py namespace.

load_data

era_py.load_data(name, *, restore_categories=True)

Load a packaged dataset by name from the shared packaged data directory.

Parameters

Name	Type	Description	Default
name	str	Dataset name, e.g. “zhang_2007_windows”.	required
restore_categories	bool	Restore factor columns as pandas Categorical using sidecar metadata.	True

load_farr_rda

era_py.load_farr_rda(name, *, timeout=30.0)

Load an .rda dataset named name from the farr GitHub repo and return the object.

Parameters

Name	Type	Description	Default
name	str	Dataset/object name (and file stem), e.g. “camp_scores”.	required
timeout	float	Requests timeout in seconds.	30.0

Returns

Name	Type	Description
	The R object stored in the .rda under key name.

get_ff_daily_factors

era_py.get_ff_daily_factors(
    dataset='F-F_Research_Data_Factors_daily',
    *,
    start=None,
    end=None,
    timeout=30.0,
)

Download Ken French daily factor data as a pandas DataFrame.

get_ff_ind

era_py.get_ff_ind(ind, *, timeout=30.0)

Download Fama-French SIC industry definitions as a pandas DataFrame.

get_size_rets_monthly

era_py.get_size_rets_monthly(timeout=30.0)

Download Ken French monthly size-portfolio returns as a pandas DataFrame.

get_me_breakpoints

era_py.get_me_breakpoints(timeout=30.0)

Download Ken French market-equity breakpoints as a pandas DataFrame.

get_trading_dates

era_py.get_trading_dates(conn)

get_annc_dates

era_py.get_annc_dates(db)

get_event_dates

era_py.get_event_dates(
    data,
    conn,
    permno='permno',
    event_date='event_date',
    win_start=0,
    win_end=0,
    end_event_date=None,
)

get_event_rets

era_py.get_event_rets(
    data,
    conn,
    permno='permno',
    event_date='event_date',
    win_start=0,
    win_end=0,
    end_event_date=None,
    suffix='',
    data_dir=None,
)

load_parquet

era_py.load_parquet(con, table, schema, data_dir=None)

copy_inline

era_py.copy_inline(df, con)

wrds_connect

era_py.wrds_connect(wrds_id=None, use_local=False)

wrds_table

era_py.wrds_table(db, table, schema)

ols_dropcollinear

era_py.ols_dropcollinear(data, formula, tol=1e-10, **fit_kwargs)

Drop-in replacement for ols(data, formula, …), but with R-like rank-deficiency handling: - refits using a full-rank subset of columns - returns a normal OLSResults (so fm.model.exog works) - stores dropped terms and ‘full’ (NaN-filled) coef vectors for reporting

modelsummary

era_py.modelsummary(
    models,
    *,
    coef_omit=None,
    stars=True,
    estimate=None,
    statistic=True,
    statistic_fmt='({std_error:.3f})',
    gof_map=('nobs',),
    add_rows=None,
    output='dataframe',
)

spline_smooth

era_py.spline_smooth(data, *, x, y, df=6, n=200)

Fit a cubic regression spline y ~ s(x) and return a prediction grid for plotting.

Parameters

Name	Type	Description	Default
data	DataFrame	Input data.	required
x	str	Column names for x and y.	required
y	str	Column names for x and y.	required
df	int	Degrees of freedom for the spline basis (patsy cr()).	6
n	int	Number of grid points for the smooth curve.	200

Returns

Name	Type	Description
	DataFrame with columns [x, f"{y}_smooth"] suitable for plotnine geom_line().

rdplot

era_py.rdplot(
    data,
    *,
    y,
    x,
    cutoff,
    y_label,
    x_label,
    degree=4,
    title='RD Plot',
    bins=None,
    binselect='esmv',
    masspoints='adjust',
    use_rdrobust=True,
)