Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Interaction plots. 2 continious variables.

    Hi guys,

    I thought I would make a slick graph for my interaction effect for 2 continous variables.

    dataex :
    . dataex MsciUsa_TotReturnIndex real_month exhibit_force exhibit_force_event reduce_relations reduce_relations_event

    ----------------------- copy starting from the next line -----------------------
    Code:
    * Example generated by -dataex-. To install: ssc install dataex
    clear
    input double MsciUsa_TotReturnIndex float real_month long exhibit_force double exhibit_force_event long reduce_relations double reduce_relations_event
    1158.852 408  3  11 11  27
    1201.082 409  4   8 21  58
    1167.437 410 14  38 28  67
    1117.015 411  7  15 15  37
    1133.025 412  3   4 11  29
    1152.004 413  3   7 30  67
    1119.778 414  7  15 15  28
    1157.778 415  8  20 18  44
    1203.548 416 12  40 22  51
    1178.293 417 42 130 30  63
    1205.588 418  6  10  7  16
    1164.663 419  6  11  8  19
    1182.011 420  2   3 19  37
    1214.883 421  3   8 25  55
    1263.074 422  8  14 29  70
    1297.222 423  3   3 15  34
    1338.256 424  8  16 15  34
    1390.833 425  1   2 20  43
    1425.372 426  6  13 22  44
    1474.114 427  2   4 22  46
    1474.553 428  6  15  9  19
    1541.034 429  3   5 11  18
    1540.647 430  4   5 22  61
    1608.724 431  2   4 13  26
    1633.386 432  5   8 15  29
    1692.017 433 14  37 34  75
    1709.998 434 13  22 36  66
    1727.687 435 11  23 44 123
     1753.61 436  4   6 40  97
    1800.809 437 11  23 41  86
    1811.811 438 23  45 49  90
    1732.479 439 14  33 51 111
    1771.655 440 41  94 34  61
    1871.283 441  8  16 40  92
    1919.512 442  6   9 35  69
    2064.995 443 10  22 35  66
    2026.293 444  3   5 25  44
    2165.497 445  7  14 19  34
    2181.224 446  7  11 47 126
    2084.009 447 10  18 43  79
    2221.712 448  1   1 36  88
    2347.995 449  7  12 57 123
    2455.442 450  8  13 24  47
     2650.06 451  7  12 32  66
    2493.214 452  3   7 33  60
    2623.869 453 10  26 36  70
     2552.82 454 10  15 56 130
     2676.92 455 12  20 43  81
    2716.982 456  7  14 41 112
     2752.48 457 22  39 36  60
    2946.083 458 12  21 56 108
     3099.36 459  7  16 34  72
    3136.193 460 16  29 96 262
     3073.75 461  6  11 65 158
    3206.355 462 13  23 33  77
     3175.11 463 22  59 56 119
    2733.618 464 14  30 34  66
    2913.732 465  7  14 37  72
    3139.696 466 24  46 56 124
    3353.374 467 21  50 55 123
    3551.723 468 26  61 55 109
    3705.311 469 16  28 51  94
    3601.942 470 10  22 63 197
    3751.828 471 24  42 55 108
    3887.652 472 18  38 40  89
     3797.28 473 19  31 43  83
    4001.631 474 13  27 47 101
    3872.025 475 17  44 45  86
    3847.674 476 18  39 71 153
    3733.527 477 22  47 61 124
    3977.809 478 12  21 66 133
    4063.205 479 12  27 56 134
    4346.659 480  7  12 52 121
    4113.942 481 10  20 39  88
    4016.239 482 14  37 48 122
    4417.244 483 11  23 51 118
    4274.448 484 14  39 39  81
    4161.501 485 10  31 59 132
    4261.432 486  6  13 36  87
    4185.351 487  8  13 46  92
    4403.062 488  7  15 39  71
    4163.637 489 17  54 42 112
    4134.069 490  5   9 43  88
    3807.643 491  3   5 32  72
    3801.778 492  8  15 39  85
     3962.02 493 19  60 48 108
    3591.656 494 14  33 58 121
    3317.897 495 19  40 57 137
    3673.604 496 17  30 48 101
    3658.387 497 10  18 61 140
    3600.637 498 10  23 58 102
    3539.477 499 19  40 48  77
    3292.648 500 30  57 88 190
    3036.331 501 43  95 46  94
    3156.673 502 25  59 47 101
     3294.62 503 22  47 55 129
     3344.37 504 28  51 61 112
    3272.428 505 26  60 55 115
    3310.304 506 19  37 21  36
      3350.8 507 12  24 19  33
    end
    format %tm real_month
    ------------------ copy up to and including the previous line ------------------

    Listed 100 out of 240 observations
    Use the count() option to list more

    // add vce(unconditional) to include standard errors appropriate for population inferences:
    quietly margins, at(exhibit_force=(0(20)160) reduce_relations=(0(60)600)) saving(predictions,replace)
    use predictions, clear
    list _at1 _at2 _margin 1/5
    rename _at1 exhibit_force
    rename _at2 reduce_rela
    rename _margin msci_USA_ri
    list exhibit_force reduce_relations msci_USA_ri in 1/5, abbreviate(9)
    //now we are ready to fill in a basic twoway contour to view the results
    twoway contour msci_USA_ri exhibit_force reduce_relations
    // this makes a graph
    //however if I would like to see te changes in effect and view the predicted probability of hypertension in increments of 0.1
    twoway (contour msci_USA_ri exhibit_force reduce_relations, ccuts(0(0.1)1.0)),
    xlabel(0(20)160)
    ylabel(0(60)600, angle(horizontal))
    xtitle("exhibit_force (observations)")
    ytitle("reduce_relations (observations)")
    ztitle("Predicted_msci_USA_ReturnIndex")
    title("Predicted_msci_USA_ReturnIndex by exhibit_force and reductions_of_relations")Predicted_msci_USA_Return Index
    error:
    . // add vce(unconditional) to include standard errors appropriate for population inferences:
    . quietly margins, at(exhibit_force=(0(20)180) reduce_relations=(0(60)600)) saving(predictions,replace)
    variable 'reduce_relations' not found in list of covariates
    r(322);

    thus this error gives expected outputs for reduce_relations.

    unsure how to proceed.
    exhibit_force ranges from 1 to 161
    reduce relation from 4 to 606.

    Q:
    - how to get stata to found reduce_relations in my list of covariates
    - are the effects of omiting observations for 161 exhibting force and 601 602 603 604 605 in reduce_relations negligible, if goal is to show presence interaction effect.

    thank you

  • #2
    also dataex (your list of variables here ) is very cool.

    Comment


    • #3
      #solved

      Comment

      Working...
      X