Confusion in PPM output

[artemshcherbak25]

Hello! We noticed that the PPM output is in decreasing order of history length 12, 8, 4. The reason for this is to initialize the history_lengths array in the mica.h file:

const UINT32 history_lengths[NUM_HIST_LENGTHS] = {12,8,4};

And the captions to the output data are reverse, in ascending order of the length of the story. See tableGen.sh file. and README:

GAg_mispred_cnt_4bits PAg_mispred_cnt_4bits GAs_mispred_cnt_4bits PAs_mispred_cnt_4bits GAg_mispred_cnt_8bits PAg_mispred_cnt_8bits GAs_mispred_cnt_8bits PAs_mispred_cnt_8bits GAg_mispred_cnt_12bits PAg_mispred_cnt_12bits GAs_mispred_cnt_12bits PAs_mispred_cnt_12bits

If we're right, can we change the history_length array from {12,8,4} to {4,8,12}?

Artem

[boegel]

I haven't used MICA for a very long time, and I don't have the time to deep-dive back into it unfortunately.

I'm hoping that @amirjamez can be of help here...

[amirjamez]

Hi,

Sure, you can. If you are using the tableGen.sh to generate a dataset corresponding to the type of analysis MICA generates, feel free to match the header with the actual analysis.

Also, feel free to push an MR for the change so I take a look at it.

• Amir

On Thu, Feb 1, 2024, 03:20 Kenneth Hoste @.***> wrote:

I haven't used MICA for a very long time, and I don't have the time to deep-dive back into it unfortunately.

I'm hoping that @amirjamez https://github.com/amirjamez can be of help here...

— Reply to this email directly, view it on GitHub https://github.com/boegel/MICA/issues/26#issuecomment-1921093795, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACECKHRTFPCR7ALCI7Y3QETYRN3AXAVCNFSM6AAAAABCUPT5MKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMRRGA4TGNZZGU . You are receiving this because you were mentioned.Message ID: @.***>

[artemshcherbak25]

Hi, Sure, you can. If you are using the tableGen.sh to generate a dataset corresponding to the type of analysis MICA generates, feel free to match the header with the actual analysis. Also, feel free to push an MR for the change so I take a look at it. - Amir … On Thu, Feb 1, 2024, 03:20 Kenneth Hoste @.> wrote: I haven't used MICA for a very long time, and I don't have the time to deep-dive back into it unfortunately. I'm hoping that @amirjamez https://github.com/amirjamez can be of help here... — Reply to this email directly, view it on GitHub <#26 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACECKHRTFPCR7ALCI7Y3QETYRN3AXAVCNFSM6AAAAABCUPT5MKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMRRGA4TGNZZGU . You are receiving this because you were mentioned.Message ID: @.>

Thanks for the reply!

We create a pull request for changes related with the issue - https://github.com/boegel/MICA/pull/27

BR, Artem