how to check the influence of binary variable on continues variable that changes in time

[Lili23]

Hi,

I have a dataset concerning cryptocurrencies rates in time. There are 10 types of currencies. I added a several binary variables to it and I would like to check it's influence on the price. In other words, if a given variable causes that this currency price changes/increases quicker that those that do not have this feature. I paste the data sample:

label Currency="Currency" Date="Date" Open="Open" High="High" Low="Low" Close="Close" Volume="Volume" Market Cap="Market Cap" Deflationary="Deflationary" Mining_coin="Mining_coin" Algorythm_SHA_256="Algorythm_SHA_256" Algorythm_Scrypt="Algorythm_Scrypt" Al
gorytm_Ethash="Algorytm_Ethash" Proof_of_work="Proof_of_work" Proof_of_stake="Proof_of_stake" Proof_of_revenue="Proof_of_revenue" More_than_three_letters_in_short="More_than_three_letters_in_shortcut";
datalines;
ripple 24-Apr-19 0.321114 0.321282 0.296982 0.302318 1,517,791,002 . . 1 0 0 0 0 0 0 0 0
ripple 23-Apr-19 0.323844 0.328396 0.320919 0.321222 1,077,333,990 . . 1 0 0 0 0 0 0 0 0
ripple 22-Apr-19 0.322277 0.32935 0.320237 0.323934 1,131,094,080 . . 1 0 0 0 0 0 0 0 0
ripple 21-Apr-19 0.328678 0.329627 0.318746 0.322449 1,005,803,846 . . 1 0 0 0 0 0 0 0 0
ripple 20-Apr-19 0.331871 0.333213 0.324969 0.328476 931,570,799 . . 1 0 0 0 0 0 0 0 0

I tried logistic regression but it does not measure the influence on price. I also tested the correlation but I'm not sure if I can somehow use it for this purpose. What kind of analysis would you recommend for such test? Also, the data is left-censored.

Thank you!

[PaigeMiller]

It sounds like you are saying that the independent (predictor) variable is the binary variable, and the dependent (response) variable is the price.

Is that a correct understanding?

[Lili23]

Yes, that's correct.

[PaigeMiller]

---

@Lili23 wrote:
Yes, that's correct.

---

This is an ANOVA, which you can run in PROC GLM, with the binary variables in the CLASS statement. However, this does not account for the left-censoring of the response, which is something I have never really had to deal with. I believe that left censored data can be modeled via a Tobit analysis; and I leave it up to you to do the searching to find out how to do that in SAS.

[Ksharp]

Check EFFECTPLOT statement of proc logistic .

@Rick_SAS  wrote a blog about it before .