In [1]:

import sys
sys.path.append('../')
import pandas as pd
import numpy as np
import forestplot as fp
pd.set_option('display.max_rows', 100)

df = (pd.read_csv('../examples/data/sleep-mmodel.csv')
      .query("model=='all' | model=='young kids'")
     )
df.head(3)

Out[1]:

	var	coef	se	T	pval	r2	adj_r2	ll	hl	model	group	label
0	age	0.994889	1.969249	0.505213	0.613625	0.127289	0.103656	-2.873825	4.863603	all	age	in years
3	age	22.634017	15.495320	1.460700	0.149315	0.178147	-0.013619	-8.361238	53.629272	young kids	age	in years
4	black	-84.796612	82.150125	-1.032215	0.302454	0.127289	0.103656	-246.185715	76.592491	all	other factors	=1 if black

In [2]:

std_opts = dict(
    dataframe = df,
    estimate = "coef",
    ll  ="ll", 
    hl = "hl",
    varlabel = "var",
    model_col = "model",
    return_df=True,
)

Normie usage¶

In [3]:

# Vanilla
_df, ax = fp.mforestplot(**std_opts)

In [4]:

# Vanilla
# Color rows and add borders
_df, ax = fp.mforestplot(**std_opts, color_alt_rows=True, despine=False)

Make table¶

In [5]:

# Vanilla
# Color rows
# Add annotation and make it a table
# Add right annotations
_df, ax = fp.mforestplot(**std_opts, color_alt_rows=True, 
                         annote=["label"], annoteheaders=["Description"], 
                         rightannote=["group"],
                         right_annoteheaders=["Group"],
                         table=True)

Variable Groups¶

In [6]:

std_opts = dict(
    dataframe = df,
    estimate = "coef",
    ll  ="ll", 
    hl = "hl",
    varlabel = "label",
    model_col = "model",
    return_df=True,
)

In [7]:

# With groups
_df, ax = fp.mforestplot(**std_opts, color_alt_rows=True, 
                         groupvar="group",
                         rightannote=["group"],
                         right_annoteheaders=["Group"],
                         variable_header="Covariate",
                         table=True)

In [8]:

# With groups and group ordering
group_order = ["family factors", "health factors", "labor factors", "age", "occupation", "area of residence", "other factors"]
_df, ax = fp.mforestplot(**std_opts, color_alt_rows=True, 
                         groupvar="group",
                         group_order=group_order,
                         rightannote=["group"],
                         right_annoteheaders=["Group"],
                         variable_header="Covariate",
                         table=True)

In [9]:

# With groups and group ordering
group_order = ["family factors", "health factors", "labor factors", "age", "occupation", "area of residence", "other factors"]
_df, ax = fp.mforestplot(**std_opts, color_alt_rows=True, 
                         groupvar="group",
                         group_order=group_order,
                         rightannote=["group"],
                         right_annoteheaders=["Group"],
                         variable_header="Covariate",
                         capitalize="capitalize",                         
                         table=True)

In [10]:

# Add groups
_df, ax = fp.mforestplot(**std_opts,
                      groupvar='group',
                     )

In [11]:

# Add groups
# With border + color alternate rows
_df, ax = fp.mforestplot(**std_opts,
                         groupvar='group',
                         despine=False,
                         color_alt_rows=True,
                        )

In [12]:

# Add groups
# Add annotation on right
_df, ax = fp.mforestplot(**std_opts,
                      groupvar='group',
                      rightannote=['group'],
                      right_annoteheaders=['Variable group'],
                     )

More annotations on right¶

In [13]:

std_opts = dict(
    dataframe = df,
    estimate = "coef",
    ll  ="ll", 
    hl = "hl",
    varlabel = "var",
    model_col = "model",
    rightannote=['label','group'],
    right_annoteheaders=['Description', 'Group'],
    return_df=True,
)

In [14]:

# No groups
# Add multiple annotation on right
_df, ax = fp.mforestplot(**std_opts,
                     )

In [15]:

# No groups
# Add annotation on right
# No flush
_df, ax = fp.mforestplot(**std_opts,
                      flush=False,
                     )

In [16]:

# No groups
# Add annotation on right
# Add row colors
# No flush
_df, ax = fp.mforestplot(**std_opts,
                      flush=False,
                      color_alt_rows=True,
                     )

In [17]:

# W/ groups
# Add annotation on right
# Add row colors
# No flush
std_opts = dict(
    dataframe = df,
    estimate = "coef",
    ll  ="ll", 
    hl = "hl",
    varlabel = "label",
    model_col = "model",
    rightannote=['var','group'],
    right_annoteheaders=['Field', 'Group'],
    return_df=True,
)
_df, ax = fp.mforestplot(**std_opts,
                      groupvar='group',
                      flush=False,
                      color_alt_rows=True,
                     )

More annotations on left¶

In [18]:

std_opts = dict(
    dataframe = df,
    estimate = "coef",
    ll  ="ll", 
    hl = "hl",
    varlabel = "label",
    model_col = "model",
    rightannote=['var','group'],
    right_annoteheaders=['Field', 'Group'],
    annote=['var','group'],
    annoteheaders=['Field', 'Group'],      
    return_df=True,
)

In [19]:

# No groups
_df, ax = fp.mforestplot(**std_opts,
                     )

In [20]:

# Add group
_df, ax = fp.mforestplot(**std_opts,
                      groupvar='group',
                     )

In [21]:

# Add group
# Color row
_df, ax = fp.mforestplot(**std_opts,
                      groupvar='group',
                      color_alt_rows=True,
                     )

In [22]:

# Add group
# Keep borders
_df, ax = fp.mforestplot(**std_opts,
                      groupvar='group',
                      color_alt_rows=True,
                      **{'despine':False, "figsize":(4,8)}
                     )

In [23]:

# Add group
# Keep borders
# Add X-label
_df, ax = fp.mforestplot(**std_opts,
                         groupvar='group',
                         color_alt_rows=True,
                         **{'despine':False, "figsize":(4,8)}
                        )

Legends¶

In [24]:

# Add group
# Color row
# Change legend label
# Add x-label
_df, ax = fp.mforestplot(**std_opts,
                         groupvar='group',
                         modellabels=["Model 1", "Model 2"],
                         xlabel="X-label",
                         color_alt_rows=True,
                        )

In [25]:

# Add group
# Color row
# Change legend label
# Add x-label
# Legend at top
_df, ax = fp.mforestplot(**std_opts,
                         groupvar='group',
                         modellabels=["Model 1", "Model 2"],
                         xlabel="X-label",
                         color_alt_rows=True,
                         bbox_to_anchor=(.5,.95)
                        )

In [26]:

# Add group
# Color row
# Change legend label
# Add x-label
_df, ax = fp.mforestplot(**std_opts,
                         groupvar='group',
                         modellabels=["Model 1", "Model 2"],
                         xlabel="X-label",
                         color_alt_rows=True,
#                          bbox_to_anchor=None,
                        )

More models (cluttered)¶

In [27]:

df = (pd.read_csv('../examples/data/sleep-mmodel.csv')
#       .query("model=='all' | model=='young kids'")
     )

df

Out[27]:

	var	coef	se	T	pval	r2	adj_r2	ll	hl	model	group	label
0	age	0.994889	1.969249	0.505213	6.136249e-01	0.127289	0.103656	-2.873825	4.863603	all	age	in years
1	age	1.482893	2.885462	0.513919	6.077171e-01	0.180382	0.139106	-4.197237	7.163023	men	age	in years
2	age	0.953200	2.943209	0.323864	7.463431e-01	0.109105	0.053425	-4.846721	6.753121	women	age	in years
3	age	22.634017	15.495320	1.460700	1.493148e-01	0.178147	-0.013619	-8.361238	53.629272	young kids	age	in years
4	black	-84.796612	82.150125	-1.032215	3.024539e-01	0.127289	0.103656	-246.185715	76.592491	all	other factors	=1 if black
5	black	-132.967109	132.581166	-1.002911	3.167759e-01	0.180382	0.139106	-393.957639	128.023420	men	other factors	=1 if black
6	black	-68.350631	108.901754	-0.627636	5.308819e-01	0.109105	0.053425	-282.953621	146.252360	women	other factors	=1 if black
7	black	-119.741039	296.432953	-0.403940	6.876932e-01	0.178147	-0.013619	-712.695230	473.213151	young kids	other factors	=1 if black
8	clerical	22.505827	48.963263	0.459647	6.459627e-01	0.127289	0.103656	-73.685593	118.697248	all	occupation	=1 if clerical worker
9	clerical	-229.275962	102.888221	-2.228399	2.665377e-02	0.180382	0.139106	-431.814921	-26.737003	men	occupation	=1 if clerical worker
10	clerical	106.507258	61.127321	1.742384	8.281375e-02	0.109105	0.053425	-13.950911	226.965427	women	occupation	=1 if clerical worker
11	clerical	-239.902574	211.760200	-1.132897	2.617642e-01	0.178147	-0.013619	-663.486040	183.680893	young kids	occupation	=1 if clerical worker
12	construc	113.929558	105.045137	1.084577	2.786145e-01	0.127289	0.103656	-92.438240	320.297357	all	occupation	=1 if construction worker
13	construc	65.696059	107.889208	0.608922	5.430737e-01	0.180382	0.139106	-146.687512	278.079631	men	occupation	=1 if construction worker
14	construc	361.756430	428.264834	0.844703	3.991782e-01	0.109105	0.053425	-482.186946	1205.699805	women	occupation	=1 if construction worker
15	construc	-186.415201	358.223501	-0.520388	6.047082e-01	0.178147	-0.013619	-902.968890	530.138488	young kids	occupation	=1 if construction worker
16	educ	-7.210558	7.516736	-0.959267	3.378725e-01	0.127289	0.103656	-21.977660	7.556543	all	labor factors	years of schooling
17	educ	-7.545420	9.720341	-0.776250	4.382610e-01	0.180382	0.139106	-26.680241	11.589402	men	labor factors	years of schooling
18	educ	-6.278397	11.869883	-0.528935	5.973743e-01	0.109105	0.053425	-29.669318	17.112525	women	labor factors	years of schooling
19	educ	-20.972116	26.974032	-0.777493	4.399230e-01	0.178147	-0.013619	-74.928212	32.983981	young kids	labor factors	years of schooling
20	gdhlth	-82.580434	58.204717	-1.418793	1.565620e-01	0.127289	0.103656	-196.927272	31.766405	all	health factors	=1 if in good or excel. health
21	gdhlth	-188.284508	80.476539	-2.339620	2.001135e-02	0.180382	0.139106	-346.705309	-29.863707	men	health factors	=1 if in good or excel. health
22	gdhlth	-24.491583	87.716867	-0.279212	7.803399e-01	0.109105	0.053425	-197.347402	148.364236	women	health factors	=1 if in good or excel. health
23	gdhlth	-141.814976	203.213022	-0.697864	4.879582e-01	0.178147	-0.013619	-548.301540	264.671589	young kids	health factors	=1 if in good or excel. health
24	lhrwage	21.605697	31.989301	0.675404	4.997212e-01	0.127289	0.103656	-41.239303	84.450696	all	labor factors	log hourly wage
25	lhrwage	3.544752	47.045923	0.075347	9.399931e-01	0.180382	0.139106	-89.066745	96.156249	men	labor factors	log hourly wage
26	lhrwage	-22.580482	53.782320	-0.419850	6.749978e-01	0.109105	0.053425	-128.564512	83.403547	women	labor factors	log hourly wage
27	lhrwage	35.609317	122.323738	0.291107	7.719733e-01	0.178147	-0.013619	-209.074589	280.293223	young kids	labor factors	log hourly wage
28	prot	-9.249430	39.771462	-0.232564	8.161916e-01	0.127289	0.103656	-87.382976	68.884117	all	other factors	=1 if Protestant
29	prot	-20.967328	52.201060	-0.401665	6.882393e-01	0.180382	0.139106	-123.726888	81.792233	men	other factors	=1 if Protestant
30	prot	1.066643	61.898047	0.017232	9.862667e-01	0.109105	0.053425	-120.910327	123.043614	women	other factors	=1 if Protestant
31	prot	-24.173124	127.566497	-0.189494	8.503450e-01	0.178147	-0.013619	-279.344111	230.997863	young kids	other factors	=1 if Protestant
32	selfe	-21.274179	63.303918	-0.336064	7.369588e-01	0.127289	0.103656	-145.638720	103.090363	all	labor factors	=1 if self employed
33	selfe	3.484094	77.585707	0.044906	9.642142e-01	0.180382	0.139106	-149.246005	156.214193	men	labor factors	=1 if self employed
34	selfe	-90.685936	112.556497	-0.805693	4.212746e-01	0.109105	0.053425	-312.491002	131.119130	women	labor factors	=1 if self employed
35	selfe	-370.956512	241.153473	-1.538259	1.292427e-01	0.178147	-0.013619	-853.335278	111.422255	young kids	labor factors	=1 if self employed
36	smsa	-40.653557	39.507723	-1.029003	3.039595e-01	0.127289	0.103656	-118.268971	36.961857	all	area of residence	=1 if live in smsa
37	smsa	-27.631427	52.418593	-0.527130	5.985235e-01	0.180382	0.139106	-130.819208	75.556354	men	area of residence	=1 if live in smsa
38	smsa	-57.146400	61.521030	-0.928892	3.539449e-01	0.109105	0.053425	-178.380417	64.087616	women	area of residence	=1 if live in smsa
39	smsa	43.429572	138.683089	0.313157	7.552474e-01	0.178147	-0.013619	-233.977910	320.837053	young kids	area of residence	=1 if live in smsa
40	south	82.486980	46.377571	1.778596	7.589350e-02	0.127289	0.103656	-8.624685	173.598644	all	area of residence	=1 if live in south
41	south	79.401694	64.439832	1.232183	2.189217e-01	0.180382	0.139106	-47.450305	206.253693	men	area of residence	=1 if live in south
42	south	114.622934	69.300494	1.653999	9.952837e-02	0.109105	0.053425	-21.941379	251.187246	women	area of residence	=1 if live in south
43	south	71.851381	136.010112	0.528280	5.992538e-01	0.178147	-0.013619	-200.209350	343.912111	young kids	area of residence	=1 if live in south
44	spsepay	-0.001005	0.002648	-0.379739	7.042954e-01	0.127289	0.103656	-0.006207	0.004196	all	other factors	spousal wage income
45	spsepay	0.003069	0.004924	0.623344	5.335697e-01	0.180382	0.139106	-0.006624	0.012762	men	other factors	spousal wage income
46	spsepay	-0.000508	0.003621	-0.140324	8.885302e-01	0.109105	0.053425	-0.007644	0.006628	women	other factors	spousal wage income
47	spsepay	-0.004701	0.011443	-0.410849	6.826471e-01	0.178147	-0.013619	-0.027591	0.018188	young kids	other factors	spousal wage income
48	totwrk	-0.151922	0.020366	-7.459430	3.698705e-13	0.127289	0.103656	-0.191934	-0.111911	all	labor factors	mins worked per week
49	totwrk	-0.191008	0.029765	-6.417152	5.964764e-10	0.180382	0.139106	-0.249602	-0.132414	men	labor factors	mins worked per week
50	totwrk	-0.125512	0.032713	-3.836783	1.621349e-04	0.109105	0.053425	-0.189977	-0.061048	women	labor factors	mins worked per week
51	totwrk	-0.112995	0.076363	-1.479712	1.441818e-01	0.178147	-0.013619	-0.265743	0.039753	young kids	labor factors	mins worked per week
52	yrsmarr	-0.075958	2.009356	-0.037802	9.698602e-01	0.127289	0.103656	-4.023464	3.871548	all	family factors	years married
53	yrsmarr	0.224116	2.934791	0.076365	9.391834e-01	0.180382	0.139106	-5.553120	6.001352	men	family factors	years married
54	yrsmarr	-1.414137	2.931985	-0.482314	6.300539e-01	0.109105	0.053425	-7.191939	4.363665	women	family factors	years married
55	yrsmarr	-31.211144	16.914011	-1.845283	6.993267e-02	0.178147	-0.013619	-65.044204	2.621915	young kids	family factors	years married

In [28]:

std_opts = dict(
    dataframe = df,
    estimate = "coef",
    ll  ="ll", 
    hl = "hl",
    varlabel = "label",
    model_col = "model",
    rightannote=['var','group'],
    right_annoteheaders=['Field', 'Group'],
    annote=['var','group'],
    annoteheaders=['Field', 'Group'],      
    return_df=True,
)

In [29]:

# Add group
# Color row
# Change legend label
# Add x-label
_df, ax = fp.mforestplot(**std_opts,
                         groupvar='group',
                         xlabel="X-label",
                         color_alt_rows=True,
#                          bbox_to_anchor=None,
                        )

In [30]:

# Add group
# Color row
# Change legend label
# Add x-label
modellabels=["Have young kids", "Men", "Women", "Full sample"]
_df, ax = fp.mforestplot(**std_opts,
                         groupvar='group',
                         modellabels=modellabels,
                         xlabel="X-label",
                         color_alt_rows=True,
                        )

In [31]:

# Add group
# Color row
# Change legend label
# Add x-label
# Legend at top
modellabels=["Have young kids", "Women", "Men", "Full sample"]
_df, ax = fp.mforestplot(**std_opts,
                         groupvar='group',
                         modellabels=modellabels,
                         xlabel="X-label",
                         color_alt_rows=True,
                         bbox_to_anchor=(.5,.95)
                        )

In [32]:

# Add group
# Color row
# Change legend label
# Add x-label
# Legend at top
modellabels=["Have young kids", "Women", "Men", "Full sample"]
_df, ax = fp.mforestplot(**std_opts,
                         modellabels=modellabels,
                         xlabel="X-label",
                         color_alt_rows=True,
                         flush=False,
                        )

Camera ready¶

In [33]:

df = (pd.read_csv('../examples/data/sleep-mmodel.csv')
      .query("model=='all' | model=='young kids'")
     )
std_opts = dict(
    dataframe = df,
    estimate = "coef",
    ll  ="ll", 
    hl = "hl",
    varlabel="label",
    model_col="model",
)
df.head()

Out[33]:

	var	coef	se	T	pval	r2	adj_r2	ll	hl	model	group	label
0	age	0.994889	1.969249	0.505213	0.613625	0.127289	0.103656	-2.873825	4.863603	all	age	in years
3	age	22.634017	15.495320	1.460700	0.149315	0.178147	-0.013619	-8.361238	53.629272	young kids	age	in years
4	black	-84.796612	82.150125	-1.032215	0.302454	0.127289	0.103656	-246.185715	76.592491	all	other factors	=1 if black
7	black	-119.741039	296.432953	-0.403940	0.687693	0.178147	-0.013619	-712.695230	473.213151	young kids	other factors	=1 if black
8	clerical	22.505827	48.963263	0.459647	0.645963	0.127289	0.103656	-73.685593	118.697248	all	occupation	=1 if clerical worker

In [34]:

_df, ax = fp.mforestplot(**std_opts,
                         color_alt_rows=True,
                         groupvar="group",
                         table=True,
                         rightannote=["var", "group"],
                         right_annoteheaders=["Variable", "Variable group"],
                         xlabel="Coefficient (95% CI)",
                         modellabels=["Have young kids", "Full sample"],
                         mcolor=["#CC6677", "#4477AA"],
                         xticks=[-1200,-600, 0, 600],
                         return_df=True,
                         # Additional kwargs for customizations
                         **{"marker": "D",  # set maker symbol as diamond
                            "markersize": 35,  # adjust marker size
                            "xlinestyle": (0, (10, 5)),  # long dash for x-reference line 
                            "xlinecolor": "#808080",  # gray color for x-reference line
                            "xtick_size": 12,  # adjust x-ticker fontsize
                            "xlinestyle": (0, (10, 5)),  # long dash for x-reference line
                            "xlinecolor": ".8",  # gray color for x-reference line                           
                           }                           
                        )