These are synthetic teaching data. No row describes a real person, child, woman, household, village or district; names such as “District A” and “PHC Urban”, and every ID, are invented. Several are built so that the model in the book gives the numbers printed there.

How to use them

Unzip so that the folder data sits in your R working directory. The book's code then runs as printed:

cl <- read.csv("data/clinic_children.csv")

Or read a file straight from this site, without downloading it:

cl <- read.csv("https://drharshmaheshwari.com/doctors/books/statistics-first-principles-to-regression/data/clinic_children.csv")

Empty cells are missing values (NA in R). The files have no comment lines.

The files

bp_methods.csv30 rows
Read in §16.3
3 columns
adult_id, sbp_mercury, sbp_digital
cbhi_survey.csv500 rows
Read in §7.2, §8.5, §9.3, §10.2, §10.4, §10.6, §13.2, §13.3, §15.1, §15.2, §15.3, §15.4, §15.5
12 columns
hh_id, village_id, distance_km, wealth_index, maternal_education, household_size, cbhi_member, facility_use_past_month, child_sick_2wk, sought_formal_care, reason_not_joining, satisfaction_1to5
clinic_children.csv100 rows
Read in §1.6, §7.2, §9.2, §9.3, §9.5, §11.2, §11.3, §12.2, §12.4, §12.5, §13.1
11 columns
child_id, seen_today, arrival_order, age_months, sex, weight_kg, height_cm, nutritional_status, diarrhoea_episodes_month, wealth_quintile, maternal_education
cohort_dropout.csv300 rows
Read in §14.4, §14.5, §14.6, §14.7, §14.8
7 columns
hh_id, intervention, income_k, distance_km, time_months, dropout, censor_reason
ebf_diarrhoea_cohort.csv200 rows
Read in §8.7, §9.1, §10.6, §11.7, §17.3
4 columns
infant_id, phc, not_ebf, diarrhoea
journey_mlr.csv240 rows
Read in §12.4, §12.6
5 columns
child_id, age_months, sex, wealth_index, weight_kg
kappa_anaemia.csv100 rows
Read in §16.3
3 columns
patient_id, rater_A, rater_B
km_eight.csv8 rows
Read in §14.4
3 columns
member_id, time_months, dropout
lifetable_district.csv7 rows
Read in §7.5
4 columns
age_start, age_end, midyear_pop, deaths
malaria_villages.csv10 rows
Read in §13.5
4 columns
village, near_water, population, cases
matched_casecontrol.csv160 rows
Read in §10.8, §13.2
3 columns
pair_id, case, not_ebf
meta_three_trials.csv3 rows
Read in §15.7
5 columns
study, events_int, n_int, events_ctl, n_ctl
nfhs_extract.csv640 rows
Read in §15.2
9 columns
woman_id, stratum, district, residence, psu, weight, age, education, anaemic
paired_hb_ifa.csv10 rows
Read in §9.2
3 columns
id, hb_before, hb_after
pps_villages.csv10 rows
§6.2.6 PPS worked list
2 columns
village, population
sign_test_hb.csv12 rows
Read in §10.1, §10.3
3 columns
woman_id, hb_before, hb_after
standardisation_districts.csv6 rows
Read in §7.4
7 columns
age_group, popA, deathsA, popB, deathsB, std_pop, std_rate_per1000

How they were built

make_data.R is the script that wrote them, with a fixed random seed for each dataset, so the files can be rebuilt exactly (about three minutes in R; it needs the survival package). Its header explains how the engineered datasets were made. One file, paired_hb_ifa.csv, is a ten-row table typed by hand.

Licence

The same as the book: CC BY-NC-SA 4.0. Free to copy, share and adapt for non-commercial use, with credit, under the same licence. If they help you, you can support this project.

Read the book online