I have two time series x and y which roughly cover the same period of time. The data is in daily form however there are some days that have data in one dataset but no data in the other. I wish to use matlab to create two data-sets of equal size with matching dates. Essentially I wish to remove the days that don't have data in both x and y. Is there a simple way to do this? Thanks.
You could use an inner join see help join if you are able to convert your timeseries into datasets. If not you could use the ismember function, but this time you should do it only on the dates.
Something like this will work:
a = {'2015-01-01', '2015-02-02', '2015-03-03'};
b = {'2015-01-01', '2015-03-03', '2015-04-04'};
newA = a(ismember(a,b));
newB = b(ismember(b,a));
Related
I want to compare how different campaigns are progressing based on number of days into the campaign rather than by date (see day1, day2, etc... on the x-axis below).
Here is my DAX code, but I can't get it to work. Any help would be much appreciated...
**Normalised Campaign Metrics =
VAR DateReached = CALCULATE(MIN(Days[Day]),db[PAYMENT_DATE]<> BLANK(), KEEPFILTERS(db[PRODUCT_CODE SWITCH]))
VAR MaxDate = CALCULATE(MAX(db[PAYMENT_DATE]),KEEPFILTERS(db[PRODUCT_CODE SWITCH]))
VAR DayNo = SELECTEDVALUE(Days[Day])
RETURN CALCULATE(count(db[PAYMENT_DATE]),
FILTER(ALL(db[PAYMENT_DATE]),
DateReached+DayNo && DateReached+DayNo<=MaxDate))**
Many thanks!
enter image description here
I would recommend solving this through manipulating your actual data rather than a complex DAX measure. If you are familiar with star schema modelling, I would solve this problem by adding a new column to your fact table that calculates how many days from the start date the payment occurred and then connect this column to a new "Days Passed" dimension that is simply a list of numbers from 1 to however many days you need. Then, you can use this new dimension as the source data for your x axis and use a standard payment amount measure for your y axis.
I recommend to create a dimension table as the relative basis to comparison with inactive relationship. Here is a video about it:
https://youtu.be/knXFVf2ipro
I'm having some issues creating a function with the following parameters:
Ndata = extperiod(data, year, month,time)
The data is a table with 3 columns, which from left to right are:
year/month/date, time, temperature
My goal is to create a function which can extract a time and a year/month, irrespective of the date and find it's corresponding temperature.
I need to avoid using for loops
I've been advised to use floor and find, where floor(YYYYMMDD/100) = YYYY*100 + MM, which I somehow want to integrate to my function.
I've previously found a way to extract all temperatures from the data for a given day, as follows:
k = find(data(:,1)==19750101);
data(k(1):k(end),3)
I'm trying to incorporate this method, but I think that the hint "floor(YYYYMMDD/100)" throws me a of a little.
I have tried with find(data(:,1)==floor(YYYYMMDD/100)), where I would think that I'd be given all dates with a specific year and month. For example:
find( data(:,1) == floor(19660101/100) )
I thought this would give me all points in the column vector where the value is 196601. But it doesn't.
What could I try differently?
From your explanation, your want to get all temperature for a given month, no matter time and day.
So you want to find dates that are comprised in the range [YYYYMM ; YYYY{MM+1}[ or [YYYYMM ; {YYYY+1}01[ in the case of selecting December.
Recall that you store the complete date in your table. So you need to apply your operator floor to both sides of your query, not only on the query value, because no date is floor(YYYYMMDD/100)!
As a result, try the following:
find( floor(data(:,1)/100) == floor(19660101/100) )
So I have this cute pretty function which requires two input: price vector and date vector:
function [tsve, timeint, dbin,adjtsve, stringdate,rvavg,rv ] = TSVE( price,datetime )
Now I have the vectors price and datetime covering a long period let say 20days and I want to compute the previous function over each day. The number of observations for each day is random but I can easily manage to find the last observation of the day.
date=floor(datetime);
[a, b, c]=unique(date,'legacy'); <----This gives me the index for the last obs of the day
[~, nbin]=histc(date,a,1); <----I was thinking about accumarray so I made this, but I am not sure I can use it on 2 values simultaneously
So I was thinking, is there a way to run a function (iteratively) on each piece of the two vectors? or should I split my two vectors into pieces and run the function on each piece?
Thank you all in advance
I have a time series of measurements taken at different depths of a water column. I have divided these into individual cells (for later) and require some help on how to complete the following: e.g.
time = [733774,733774,733775,733775,733775,733776,733776];
bthD = [20,10,0,15,10,20,10];
bthA = (1000:100:1600);
%Hypsographic
Hypso = [(10:1:20)',(1000:100:2000)'];
d = [1,1.3,1,2.5,2.5,1,1.2];
data = horzcat(time',bthD',d');
uniqueTimes = unique(time);
counts = hist(time,uniqueTimes);
newData = mat2cell(data,counts,length(uniqueTimes));
So, in newData I have three cells, that correspond to different days of measurements, in each cell I have newData(:,1) being time, newData(:,2) being depth, and newData(:,3) being the measurement. I would like to find what the area is at each depth in the cells, the area at different depths is given in the variable 'Hypso'.
How could I achieve this?
Your problem formulation is excellent! Very easy to understand what you need here. All you need is the function interp1. Use the first column of Hypso, I assume, as your depth, and the second column as the area. You can use the vectorized ability of the interp1 function to find all values in one call:
areaAtDepth = interp1(Hypso(:,1),Hypso(:,2),bthD)
areaAtDepth =
Columns 1 through 6
2000 1000 NaN 1500 1000 2000
Column 7
1000
You'll notice the Nan in the third column of the output. This is because it's associated depth, 0, is outside the range of the data, or support of the data I believe. You'll need to decide what you want to do when data is outside the range, or perhaps it never should be, so an error should be logged; it's up to you! Let me know if you have any more questions!
I want to group 2 or more financial time series objects of different length.
Specifically, I have FTS a and b, a is shorter than b, and how can I get a time series c, so that c has two sub series a and b, and all the missing records in a is filled with null values. Thank you.
Sounds like the MERGE function is what you're looking for.