Install
visidata is developed with Python and ncurse libraries. Install these libraries before install Python:
sudo apt install libncurses5 libncurses5-dev libncursesw5 libncursesw5-dev
asdf install python 3.8.1
asdf global python 3.8.1
pip install visidata
asdf reshim python
vd aaa.csv
Frequently Used Operations
Create & save a new sheet
Create a new sheet with 5 column: A5
Add a new row: a
Edit the cell: e
, 123
Edit the column name: ^
Add a new column: za
Remove a row: d
Save the sheet: Ctrl-s.
Sheet operation
S: list sheets q: quit sheet gq: quit visidata
Row operation
t: toggle current row selection gt: select/unselect all row
Sampling
R <5>: take <5> samples from origin sheet
Filter row
|: text match z|: Python expression match
": copy selected rows to a new sheet
Column operation
!: Pin current column as key column
H/L: move current column one position to the left/right
-: hide the current column
C: column summary
F: histogram for the current column, add aggregator with +
Set type of column
#
: integer
%: float
@: date
~: text
Data summary
I: summary of all columns
z+: max/min/mean/sum ... of current column
Groupby
- 标记汇总特征的数据类型:这里是实数,用
%
标记; - 指定汇总方法:
+
后加上函数名称max
并回车; - (可选)标记分组特征的数据类型:由于字符串类型是默认类型,可以不标记;
- 分组汇总:按
F
键计算出新的分组汇总表; - (可选)按一定的规则对汇总结果排序:这里对汇总特征 max_avg_mpg 进行从小到大排序。