TS Filler, how to initiate a start date, replace NULL with STRING in row-1

[yingxin]

orking on cleansing time series data, i know TS Filler can fill the gap. But if there any way to generate a start date other than the given date in the dataset? For example TS should start on 2011-1-3, but the first date given is 2011-1-31, and TS filler fills missing date after 2011-1-31(interval="week"). Generate Row tool may help?

there are a column of char/string data, how can i fill the NULL with the same STRING from previuos row?

test data attached.

Thanks!

[s_pichaipillai]

Ying,

I already responded here to get start date or end date etc. you can refer it how to use  Generate Row tool

http://community.alteryx.com/t5/Data-Preparation-Blending/DateTimeDiff/m-p/3775#M585

[s_pichaipillai]

Look at this one too

[s_pichaipillai · Accepted solution]

Ying,

Sorry, i may be confused

See another one attached

1. Convert the Date field to date using datetimeparse tool 

2. find the start date of given date from the file ---DateTimeTrim([DateTime_Out],'firstofmonth')

3. Use TS Filler tool 

4. For the rows came as null use Multi Row Formula tool to fill with the previous row

hope i am in right direction

[yingxin]

Thanks. Although in my case the Start_date is "firstMondayofMonth", your workflow provided me great ideas.

I used text_input to initiate the Start_Date, then union into the file, after that it's very close to yours. i was able to run in batch macro to process a few of such TS files, lucky for me, they all started with the same Start_Date in the text_input.

one thing I noticed about the tool, "multiple rows", if there are a few (>2) rows of NULL both at the beginning, and in the end, then "multiple rows" probably won't work, is that right?

[s_pichaipillai]

Ying

Glad to that , it Helped

For your Multirow Formula Question,

TRY to change the "Values for Rows that Dont Exit" to set values to Closest