The RawQuery command sends the string query to one of two models from OpenAI: ChatGPT or GPT-4, through a Maplesoft server; and returns the response as a string. All other commands in the NaturalLanguage package are routed through RawQuery, with some pre- and/or post-processing.

Maple can interface with OpenAI's GPT-4 and ChatGPT models. The default model is GPT-4; you can select ChatGPT by using the $\mathrm{model}=\mathrm{ChatGPT}$ option, or force use of the GPT-4 model by using the $\mathrm{model}=\mathrm{GPT4}$ option. Note that OpenAI may deprecate and disable models, so the set of models supported may change in the future.

Note: Large language models often generate inaccurate statements. Please keep this in mind: this is not technology for building a bridge with.

$\mathrm{with}\left(\mathrm{NaturalLanguage}\right)\:$

$\mathrm{RawQuery}\left(''plot\; sin(x)\; in\; maple''\right)$

$''To\; plot\; `sin(x)`\; in\; Maple,\; use\; the\; following\; command:\; `plot(sin(x),\; x\; =\; -Pi\; ..\; Pi);`\; Enter\; this\; command\; in\; your\; Maple\; IDE\; and\; run.\; This\; will\; plot\; a\; graph\; of\; the\; sine\; function\; from\; x\; =\; -Pi\; to\; Pi.\; If\; you\; want\; to\; change\; the\; range,\; you\; can\; modify\; the\; values\; after\; `x\; =`.''$

$\mathrm{RawQuery}\left(''maple\; code\; to\; integrate\; (x^2-1)/(x+2)''\,\mathrm{model}=\mathrm{ChatGPT}\right)$

$''To\; integrate\; the\; expression\; (x^2\; -\; 1)/(x\; +\; 2)\; using\; Maple,\; you\; can\; use\; the\; `int`\; command.\; Here\; is\; the\; Maple\; code:\; ```maple\; int((x^2\; -\; 1)/(x\; +\; 2),\; x)\; ```\; This\; will\; compute\; the\; indefinite\; integral\; of\; (x^2\; -\; 1)/(x\; +\; 2)\; with\; respect\; to\; x.''$

$\mathrm{RawQuery}\left(''Mary\; has\; seven\; apples.\; She\; gives\; two\; to\; Bob\; and\; three\; to\; Sue.\; How\; many\; does\; she\; have\; left?\; Show\; the\; steps''\right)$

$\mathrm{RawQuery}\left(''What\; is\; the\; area\; of\; a\; triangle\; as\; a\; Maple\; equation''\,\mathrm{model}=\mathrm{ChatGPT}\right)$

$''To\; find\; the\; area\; of\; a\; triangle\; using\; Maple,\; you\; can\; use\; the\; formula:\; Area\; =\; (1/2)\; *\; base\; *\; height\; Assuming\; you\; have\; the\; values\; for\; the\; base\; and\; height\; of\; the\; triangle,\; you\; can\; define\; a\; Maple\; equation\; like\; this:\; ```maple\; base\; :=\; 5;\; \#\; replace\; with\; your\; value\; for\; the\; base\; of\; the\; triangle\; height\; :=\; 8;\; \#\; replace\; with\; your\; value\; for\; the\; height\; of\; the\; triangle\; Area\; :=\; (1/2)\; *\; base\; *\; height;\; Area;\; ```\; This\; will\; return\; the\; value\; of\; the\; area\; of\; the\; triangle\; based\; on\; the\; given\; base\; and\; height.''$

$\mathrm{RawQuery}\left(''How\; do\; I\; do\; extreme\; value\; analysis?''\,\mathrm{model}=\mathrm{ChatGPT}\right)$

$Extreme value analysis (EVA) is a statistical technique used to analyze and make predictions about the extreme or rare events in a dataset. It is commonly applied in areas such as finance, insurance, engineering, and environmental science. Here are the general steps to perform extreme value analysis: 1. Data collection: Gather the historical data on the variable of interest, such as maximum or minimum values, over a specific time period. Ensure that the data is accurate, complete, and representative of the events being analyzed. 2. Data preprocessing: Clean the data by removing any outliers or erroneous values that could affect the analysis. Check for missing data and decide on a suitable approach for handling it (e.g., imputing missing values or excluding incomplete data points). 3. Distribution fitting: Identify an appropriate probability distribution that can describe the extreme values in your dataset. Commonly used distributions include the Generalized Extreme Value (GEV) distribution, the Generalized Pareto Distribution (GPD), or the Gumbel distribution. Fit the chosen distribution to your data using statistical techniques such as maximum likelihood estimation. 4. Statistical inference: Calculate the parameters of the chosen distribution (e.g., shape, location, and scale parameters) that best fit your data. Assess the goodness-of-fit of the distribution by comparing the observed data with the fitted distribution using statistical tests or visual inspections. 5. Estimation of return levels: Once the distribution is fitted, you can estimate the return levels for certain probabilities or recurrence intervals. Return levels are the values of extreme events that are expected to be exceeded, on average, once within a given time period. For example, you could estimate the return level for a 100-year flood event or a 10-year high temperature event. 6. Uncertainty assessment: Quantify the uncertainty associated with the estimated return levels using methods like confidence intervals or simulation techniques (e.g., Monte Carlo simulation). 7. Model validation: Evaluate the effectiveness of your extreme value model by testing its performance on a validation dataset that was not used in the model fitting process. Check if the model predictions align with the observed extreme events in the validation dataset. Remember that extreme value analysis assumes stationarity, which means that the statistical properties of extreme events remain consistent over time. If stationarity assumptions are violated, other methods like time series analysis or non-stationary extreme value analysis may be more appropriate.$

$\mathrm{RawQuery}\left(''What\text{'}s\; the\; equation\; for\; a\; photovoltaic\; diode?''\right)$

$''The\; equation\; for\; a\; photovoltaic\; diode,\; often\; known\; as\; solar\; cell,\; is\; given\; by\; the\; Shockley\; diode\; equation:\; I\; =\; Iph\; -\; I0\; [exp(qV/nkT)\; -\; 1]\; Where:\; I\; is\; the\; diode\; current,\; Iph\; is\; the\; photocurrent,\; I0\; is\; the\; diode\; saturation\; current,\; q\; is\; the\; electron\; charge,\; V\; is\; the\; voltage\; across\; the\; diode,\; n\; is\; the\; ideality\; factor\; (typically\; between\; 1\; and\; 2),\; k\; is\; the\; Boltzmann\; constant,\; T\; is\; the\; absolute\; temperature.''$

The OpenAI website.

The NaturalLanguage:-RawQuery command was introduced in Maple 2024.

For more information on Maple 2024 changes, see Updates in Maple 2024.