Data description

Resources¶

Introduction¶

Wikipedia: Exploratory data analysis

Useful Packages¶

R

Sample description¶

TBD

Summary of variables¶

All variables¶

R

Using skimr

skimr::skim()

Using summarytools

data_summary <- summarytools::dfSummary(
    dataset_xxx,
    varnumbers = TRUE,
    labels.col = if_label,
    graph.magnif = 1, 
    valid.col = FALSE,
    na.col = TRUE,
    style = "grid", 
    plain.ascii = FALSE,
    max.string.width = 25,
    split.table = 30,
    tmp.img.dir = "/tmp"
)
summarytools::view(
    data_summary,
    footnote = NA,
    file = file.path(path_html, paste0(name, "_summary.html"))
)

Continuous variables¶

RStata

hist(vector_name)

su <var>
su <var>, d // equivalent to `summarize, detail`
codebook <var>
histogram <var>

Categorical variables¶

RStata

table()

dplyr::count(col_name)

janitor::tabyl(col_name)

Missing values¶

RStata

finalfit::missing_plot()

misstable

Correlation analysis¶

TBD

Pearson's correlation coefficient (\(\rho\))
Rank correlation coefficients

Correlation coefficients¶

Visualization¶

Patterns across dimensions¶

TBD

Subgroup comparisons¶

Summary across waves¶

Time trends¶

Case studies¶

TBD