---
tags: hathi
---

# Hathidata overview

## Hathidata in numbers
Total number of documents: 108 395
Estimate of distinct titles: 51 196
Estimate of distinct authors: 19 481

*The real numbers are probably lower, since my cleaning method did not capture all duplicates. Sometimes the same text might also have different titles.*

### Timely distribution
Decades (last zero of each decade missing, but don't let it bother)

- earliest year: 1701
- latest year: 2011

### Authors

**Authors with the most text in the corpus**
(names cleaned and lowercased)

dickens charles: 1107
scott walter sir: 981
balzac honore de: 813
thackeray william makepeace: 769
defoe daniel: 704
cooper james fenimore: 694
lytton edward bulwer lytton: 666
trollope anthony: 616
stevenson robert louis: 574
dumas alexandre: 546

## The English subset

2730 titles, manually checked, evenly distributed over time:



*In our data: 1678. More recent files missing.*

#### Preliminary Hurst exploration

*Quite stable over decades!*

## Nobel set

Different authors: 27 (out of 118)


- the year 1800 is a mistake, will be corrected manually

An overrepresentation of Kipling:


## Documentation

About the collection: https://wiki.htrc.illinois.edu/display/COM/About+the+Collection

Tutorials: https://wiki.htrc.illinois.edu/display/COM/All+HTRC+Tutorials

Validating the Nobel workset?
https://analytics.hathitrust.org/validateworkset

HathiTrust on GitHub: https://github.com/htrc

https://github.com/tedunderwood/noveltmmeta/tree/master/metadata#4--the-manually-checked-title-subset