TWFE

from differences import simulate_data, TWFE

see plotto to configure the plots. the .plot() method takes the same arguments as the function mark_plot() in plotto

panel_data = simulate_data()  # generate data

panel_data

		y	x0	w	cat.0	cat.1	effect	cohort	intensity
entity	time
e0	1900	0.376343	0.160245	3.729827	0	0	0.0	NaN	NaN
	1901	1.664105	-1.933404	1.691201	0	0	0.0	NaN	NaN
	1902	-11.229124	-0.056813	1.786326	0	0	0.0	NaN	NaN
	1903	0.542425	0.810118	1.741609	0	0	0.0	NaN	NaN
	1904	1.111293	0.101205	0.806004	0	0	0.0	NaN	NaN
...	...	...	...	...	...	...	...	...	...
e999	1903	-12.916305	1.355023	0.688025	0	0	0.0	NaN	NaN
	1904	-7.780740	-0.381267	1.952084	0	0	0.0	NaN	NaN
	1905	-16.677634	-0.258604	1.801049	0	0	0.0	NaN	NaN
	1906	0.556390	1.504564	0.616657	0	0	0.0	NaN	NaN
	1907	-4.153630	0.472740	0.674045	0	0	0.0	NaN	NaN

8000 rows × 8 columns

twfe = TWFE(
    data=panel_data,
    cohort_name='cohort',
)

twfe.fit(
    formula='y',
    cluster_names='entity',
)

	parameter	std_error	t_stat	p_value	lower	upper
Intercept	0.170176	0.483307	0.352107	7.247582e-01	-0.777088	1.117439
-6	3.943623	1.100968	3.581961	3.410253e-04	1.785766	6.101481
-5	3.322752	1.025707	3.239476	1.197494e-03	1.312404	5.333100
-4	2.169448	0.901665	2.406047	1.612621e-02	0.402217	3.936679
-3	0.799785	0.736629	1.085737	2.775954e-01	-0.643981	2.243552
-2	2.223513	0.855707	2.598452	9.364502e-03	0.546359	3.900667
0	-0.120740	0.881288	-0.137004	8.910277e-01	-1.848032	1.606552
1	5.291089	0.708303	7.470090	8.015810e-14	3.902840	6.679338
2	12.605998	1.031534	12.220630	0.000000e+00	10.584228	14.627768
3	21.453005	0.873314	24.565050	0.000000e+00	19.741341	23.164670
4	34.619623	1.321403	26.199132	0.000000e+00	32.029720	37.209526
5	43.410681	1.207994	35.936172	0.000000e+00	41.043057	45.778306
6	54.668145	1.478442	36.976859	0.000000e+00	51.770451	57.565838
7	67.229564	1.357234	49.534260	0.000000e+00	64.569435	69.889693

twfe.plot()